Add files using upload-large-folder tool

Art_text_detection/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "Art_text_detection",
+  "modality": [
+    "image"
+  ],
+  "category": "text_detection",
+  "total": 55320,
+  "sampled": 20
+}

Art_text_detection/sampled.jsonl

@@ -0,0 +1,20 @@
+{"image": ["media/000425_gt_3034.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind all text regions in this image."}, {"from": "gpt", "value": "<ref>金港居委会消费维权联络点</ref><box>[<51><784><478><1000>]</box><ref>青浦区香花桥街道金港社区</ref><box>[<71><366><478><583>]</box><ref>网格化管理工作站</ref><box>[<118><580><396><677>]</box><ref>科普居委</ref><box>[<567><259><1000><480>]</box><ref>金港警务点</ref><box>[<637><884><958><982>]</box><ref>香花桥派出所</ref><box>[<681><804><907><870>]</box><ref>青浦区人民政府</ref><box>[<777><439><980><507>]</box><ref>二00九年三月</ref><box>[<778><480><977><542>]</box>"}]}
+{"image": ["media/006078_gt_5419_2.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all text content in the image."}, {"from": "gpt", "value": "<ref>COMPANY</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/006338_gt_1161_3.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect every text instance visible in this image."}, {"from": "gpt", "value": "<ref>三级警务区</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/006698_gt_2699_4.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all the text that appears in this image."}, {"from": "gpt", "value": "<ref>SPRING</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/010189_gt_2356_9.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all readable text."}, {"from": "gpt", "value": "<ref>of</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/010463_gt_195_3.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind all text elements in the image and provide their locations."}, {"from": "gpt", "value": "<ref>TRALN</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/014110_gt_1733_8.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all the text in this image."}, {"from": "gpt", "value": "<ref>alc.7.5%vol</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/018210_gt_4452_4.jpg"], "conversations": [{"from": "human", "value": "<image>\nScan the image for text."}, {"from": "gpt", "value": "<ref>宏伟粮油商店</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/018231_gt_3336_5.jpg"], "conversations": [{"from": "human", "value": "<image>\nScan the image and detect all readable text."}, {"from": "gpt", "value": "<ref>Custom</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/022059_gt_5385_14.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and locate every piece of text visible in this image."}, {"from": "gpt", "value": "<ref>PORK</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/022298_gt_2477_0.jpg"], "conversations": [{"from": "human", "value": "<image>\nIdentify all text regions."}, {"from": "gpt", "value": "<ref>CLINTON</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/023526_gt_1775_1.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and locate every piece of text visible in this image."}, {"from": "gpt", "value": "<ref>MING YANG</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/024898_gt_3292_0.jpg"], "conversations": [{"from": "human", "value": "<image>\nIdentify all text regions."}, {"from": "gpt", "value": "<ref>hot</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/027696_gt_4569_8.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all text in the image."}, {"from": "gpt", "value": "<ref>Musenm</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/038618_gt_4671_2.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind every text instance in this image."}, {"from": "gpt", "value": "<ref>T</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/045753_gt_1102_0.jpg"], "conversations": [{"from": "human", "value": "<image>\nIdentify all text regions present in this image.\nUse format like: <ref>car</ref><box>[<120><50><400><300>]</box><ref>person</ref><box>[<500><100><650><450>]</box>"}, {"from": "gpt", "value": "<ref>富路电动</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/049729_gt_971_1.jpg"], "conversations": [{"from": "human", "value": "<image>\nScan the image for text."}, {"from": "gpt", "value": "<ref>大碗饭煲仔饭</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/050036_gt_4707_152.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all the text in this image."}, {"from": "gpt", "value": "<ref>Distributed</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/052810_gt_4057_2.jpg"], "conversations": [{"from": "human", "value": "<image>\nScan the image for text."}, {"from": "gpt", "value": "<ref>您永远的家</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/053046_gt_4904_2.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind every text instance in this image."}, {"from": "gpt", "value": "<ref>发各种园螺母螺丝</ref><box>[<0><0><1000><1000>]</box>"}]}

Blip3_text_detection/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "Blip3_text_detection",
+  "modality": [
+    "image"
+  ],
+  "category": "text_detection",
+  "total": 294856,
+  "sampled": 20
+}

Humanref_grounding/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "Humanref_grounding",
+  "modality": [
+    "image"
+  ],
+  "category": "grounding",
+  "total": 37044,
+  "sampled": 20
+}

Humanref_grounding/sampled.jsonl

@@ -0,0 +1,20 @@
+{"image": ["media/000212_54326130.jpg"], "conversations": [{"from": "human", "value": "<image>\nGive bounding boxes for each: The two athletes who are closest to the camera and are in confrontation."}, {"from": "gpt", "value": "<ref>The two athletes who are closest to the camera and are in confrontation</ref><box>[<303><115><873><915>]</box><ref>The two athletes who are closest to the camera and are in confrontation</ref><box>[<28><133><441><951>]</box>"}]}
+{"image": ["media/006973_235583703.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every occurrence of: the people on both sides of the lady in the red vest in the middle.\nExpected output: <ref>object name</ref><box>[[x1, y1, x2, y2], [x1, y1, x2, y2], ...]</box> for each category."}, {"from": "gpt", "value": "<ref>the people on both sides of the lady in the red vest in the middle</ref><box>[<555><268><847><1000>, <71><106><380><1000>]</box>"}]}
+{"image": ["media/007310_322854542.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutline all regions of: the first person wearing gloves to the left of the first person wearing black clothes when counting from left to right."}, {"from": "gpt", "value": "<ref>the first person wearing gloves to the left of the first person wearing black clothes when counting from left to right</ref><box>[<344><134><470><841>]</box>"}]}
+{"image": ["media/010016_443816118.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn box coordinates for all: the person wearing a white suit."}, {"from": "gpt", "value": "<ref>the person wearing a white suit</ref><box>[<466><385><527><733>]</box>"}]}
+{"image": ["media/010365_533976280.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all regions matching: the person wearing khaki clothes."}, {"from": "gpt", "value": "<ref>the person wearing khaki clothes</ref><box>[<277><275><451><499>, <257><621><438><992>]</box>"}]}
+{"image": ["media/010587_592424752.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn bounding box coordinates for all: the man.\nCoordinates are [x1, y1, x2, y2], where (x1, y1) is top-left and (x2, y2) is bottom-right."}, {"from": "gpt", "value": "<ref>the man</ref><box>[<321><84><500><651>, <754><424><821><562>]</box>"}]}
+{"image": ["media/011708_884609126.jpg"], "conversations": [{"from": "human", "value": "<image>\nGive bounding boxes for each: the person sitting with crossed legs."}, {"from": "gpt", "value": "<ref>the person sitting with crossed legs</ref><box>[<85><257><279><999>]</box><ref>the person sitting with crossed legs</ref><box>[<373><504><497><1000>]</box>"}]}
+{"image": ["media/013035_1224795184.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn bounding box coordinates for all: the person wearing a gray flat cap."}, {"from": "gpt", "value": "<ref>the person wearing a gray flat cap</ref><box>[<525><159><970><760>]</box>"}]}
+{"image": ["media/017261_2312423734.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all regions matching: the person wearing a cap with a visor."}, {"from": "gpt", "value": "<ref>the person wearing a cap with a visor</ref><box>[<0><68><312><894>]</box>"}]}
+{"image": ["media/017486_2370828260.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw boxes around all: the person wearing a blue sweater."}, {"from": "gpt", "value": "<ref>the person wearing a blue sweater</ref><box>[<644><407><864><999>]</box>"}]}
+{"image": ["media/019234_2821283901.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw boxes around all: the person with long golden hair."}, {"from": "gpt", "value": "<ref>the person with long golden hair</ref><box>[<392><356><578><946>]</box>"}]}
+{"image": ["media/019558_2904036473.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn bounding box coordinates for all: the person wearing a blue-purple tie."}, {"from": "gpt", "value": "<ref>the person wearing a blue-purple tie</ref><box>[<758><184><902><837>]</box>"}]}
+{"image": ["media/022293_3608925878.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw boxes around all: the bridegroom."}, {"from": "gpt", "value": "<ref>the bridegroom</ref><box>[<244><160><620><989>]</box>"}]}
+{"image": ["media/024504_4181296290.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every: the person wearing a pink short-sleeved shirt."}, {"from": "gpt", "value": "<ref>the person wearing a pink short-sleeved shirt</ref><box>[<172><21><268><195>]</box>"}]}
+{"image": ["media/028492_508978812.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every: the person not wearing a blue hoodie."}, {"from": "gpt", "value": "<ref>the person not wearing a blue hoodie</ref><box>[<58><165><301><1000>]</box><ref>the person not wearing a blue hoodie</ref><box>[<454><204><686><1000>]</box><ref>the person not wearing a blue hoodie</ref><box>[<502><266><1000><1000>]</box>"}]}
+{"image": ["media/029735_826290922.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn bounding box coordinates for all: the person not carrying a red rectangular bag."}, {"from": "gpt", "value": "<ref>the person not carrying a red rectangular bag</ref><box>[<209><118><513><972>, <751><234><872><476>, <829><235><987><445>]</box>"}]}
+{"image": ["media/032806_182900778.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all regions matching: the first person wearing red and white clothes all over from left to right.\nExpected output: <ref>object name</ref><box>[[x1, y1, x2, y2], [x1, y1, x2, y2], ...]</box> for each category."}, {"from": "gpt", "value": "<ref>the first person wearing red and white clothes all over from left to right</ref><box>[<445><265><534><449>]</box>"}]}
+{"image": ["media/033270_302093892.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw bounding boxes around all: the person sitting on the square cabinet."}, {"from": "gpt", "value": "<ref>the person sitting on the square cabinet</ref><box>[<625><198><927><838>]</box>"}]}
+{"image": ["media/033271_302784110.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all instances of: the person wearing glasses in the middle and return their bounding boxes."}, {"from": "gpt", "value": "<ref>the person wearing glasses in the middle</ref><box>[<282><30><781><1000>]</box>"}]}
+{"image": ["media/035348_134595059.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all instances of: Miguel Cabrera and return their bounding boxes."}, {"from": "gpt", "value": "<ref>Miguel Cabrera</ref><box>[<302><31><654><1000>]</box>"}]}

Object365_dense_multi_category_detection/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "Object365_dense_multi_category_detection",
+  "modality": [
+    "image"
+  ],
+  "category": "multi_category_detection",
+  "total": 174020,
+  "sampled": 20
+}

Object365_multi_category_detection/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "Object365_multi_category_detection",
+  "modality": [
+    "image"
+  ],
+  "category": "multi_category_detection",
+  "total": 1568272,
+  "sampled": 20
+}

Openimages_groundingv2/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "Openimages_groundingv2",
+  "modality": [
+    "image"
+  ],
+  "category": "grounding",
+  "total": 1428175,
+  "sampled": 20
+}

Openimages_groundingv2/sampled.jsonl

@@ -0,0 +1,20 @@
+{"image": ["media/113605_abc764ca773ffe83.jpg"], "conversations": [{"from": "human", "value": "<image>\nProvide bounding boxes for every: A blue sippy cup with a yellow spout near the bottom center."}, {"from": "gpt", "value": "<ref>A blue sippy cup with a yellow spout near the bottom center</ref><box>[<390><906><490><1000>]</box>"}, {"from": "human", "value": "Find all bounding boxes for: A woman holding a baby in her arms."}, {"from": "gpt", "value": "<ref>A woman holding a baby in her arms</ref><box>[<75><253><998><1000>]</box>"}, {"from": "human", "value": "Give bounding boxes for each: A young girl with curly hair looking to the left."}, {"from": "gpt", "value": "<ref>A young girl with curly hair looking to the left</ref><box>[<535><520><816><1000>]</box>"}, {"from": "human", "value": "Locate all regions matching: The woman with a camera raised to her face."}, {"from": "gpt", "value": "<ref>The woman with a camera raised to her face</ref><box>[<673><321><892><605>]</box>"}, {"from": "human", "value": "Outline all regions of: A woman holding a baby in her arms."}, {"from": "gpt", "value": "<ref>A woman holding a baby in her arms</ref><box>[<75><253><998><1000>]</box>"}]}
+{"image": ["media/127113_f59d70129822ace1.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all instances of: people on the roof of the building in the background and return their bounding boxes."}, {"from": "gpt", "value": "<ref>people on the roof of the building in the background</ref><box>[<390><238><520><305>]</box>"}, {"from": "human", "value": "Draw boxes around all: crowd of people in the foreground with raised arms."}, {"from": "gpt", "value": "<ref>crowd of people in the foreground with raised arms</ref><box>[<0><547><998><997>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: bricks and windows of the large building behind the crowd."}, {"from": "gpt", "value": "<ref>bricks and windows of the large building behind the crowd</ref><box>[<131><238><819><654>]</box>"}, {"from": "human", "value": "Find all bounding boxes for: person wearing a white cap near the right edge of the image."}, {"from": "gpt", "value": "<ref>person wearing a white cap near the right edge of the image</ref><box>[<901><715><962><784>]</box>"}]}
+{"image": ["media/206397_903aa61137418824.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutline all regions of: Two people sitting on tires in front of the sign."}, {"from": "gpt", "value": "<ref>Two people sitting on tires in front of the sign</ref><box>[<348><673><600><959>, <573><685><851><953>]</box>"}, {"from": "human", "value": "Find all bounding boxes for: The large black 'OPEN 6AM-9PM' sign at the bottom of the structure."}, {"from": "gpt", "value": "<ref>The large black 'OPEN 6AM-9PM' sign at the bottom of the structure</ref><box>[<571><461><812><598>]</box>"}]}
+{"image": ["media/229153_627e25c3a80f0764.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn bounding box coordinates for all: The blue Hyundai i30 parked on the cobblestone surface."}, {"from": "gpt", "value": "<ref>The blue Hyundai i30 parked on the cobblestone surface</ref><box>[<44><321><799><876>]</box>"}, {"from": "human", "value": "Give the bounding box for each instance of: Two palm trees visible in the background on the left."}, {"from": "gpt", "value": "<ref>Two palm trees visible in the background on the left</ref><box>[<41><95><133><353>, <107><97><233><339>]</box>"}]}
+{"image": ["media/293983_94e4fb25088fea6f.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw bounding boxes around all: golden dome structure with arched opening at center."}, {"from": "gpt", "value": "<ref>golden dome structure with arched opening at center</ref><box>[<129><159><836><603>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: two vases with orange flowers on either side of the shrine."}, {"from": "gpt", "value": "<ref>two vases with orange flowers on either side of the shrine</ref><box>[<89><528><151><596>, <823><532><895><597>]</box>"}, {"from": "human", "value": "Outline all regions of: three people kneeling on the red patterned rug in the foreground."}, {"from": "gpt", "value": "<ref>three people kneeling on the red patterned rug in the foreground</ref><box>[<61><639><384><921>, <469><681><744><872>, <784><637><1000><942>]</box>"}, {"from": "human", "value": "Return box coordinates for all: blue canopy above the golden dome."}, {"from": "gpt", "value": "<ref>blue canopy above the golden dome</ref><box>[<142><0><836><75>]</box>"}]}
+{"image": ["media/384802_4fb911494c309f58.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw boxes around all: white electric guitar with black pickguard and gold hardware."}, {"from": "gpt", "value": "<ref>white electric guitar with black pickguard and gold hardware</ref><box>[<36><434><1000><1000>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: man playing guitar with star-shaped pin on shirt."}, {"from": "gpt", "value": "<ref>man playing guitar with star-shaped pin on shirt</ref><box>[<46><70><839><1000>]</box>"}, {"from": "human", "value": "Detect all instances of: vintage-style microphone on stand to the right and return their bounding boxes."}, {"from": "gpt", "value": "<ref>vintage-style microphone on stand to the right</ref><box>[<860><96><983><206>]</box>"}, {"from": "human", "value": "Find and box every occurrence of: guitar straps with metal buckle over shoulder."}, {"from": "gpt", "value": "<ref>guitar straps with metal buckle over shoulder</ref><box>[<479><342><584><659>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all instances of: person with dark hair and ear piercing holding guitar."}, {"from": "gpt", "value": "<ref>person with dark hair and ear piercing holding guitar</ref><box>[<47><70><839><1000>]</box>"}]}
+{"image": ["media/398896_ba061dcdc8985cc5.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all regions matching: Two vintage cars parked side by side on the snowy street.\nUse format like: <ref>car</ref><box>[<120><50><400><300>]</box><ref>person</ref><box>[<500><100><650><450>]</box>"}, {"from": "gpt", "value": "<ref>Two vintage cars parked side by side on the snowy street</ref><box>[<272><559><386><667>, <379><601><567><771>]</box>"}, {"from": "human", "value": "Return bounding box coordinates for all: The tall church steeple visible between the buildings in the background."}, {"from": "gpt", "value": "<ref>The tall church steeple visible between the buildings in the background</ref><box>[<200><256><248><465>]</box>"}]}
+{"image": ["media/401793_54400379a067d4e4.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw bounding boxes around all: Stack of white cassette tapes on the left side of the box."}, {"from": "gpt", "value": "<ref>Stack of white cassette tapes on the left side of the box</ref><box>[<75><540><244><936>]</box>"}, {"from": "human", "value": "Draw boxes around all: Two spiral-bound notebooks side by side at the bottom center."}, {"from": "gpt", "value": "<ref>Two spiral-bound notebooks side by side at the bottom center</ref><box>[<231><621><527><942>, <446><568><714><925>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all: Black camera with lens cap next to a book titled 'El Quijote'."}, {"from": "gpt", "value": "<ref>Black camera with lens cap next to a book titled 'El Quijote'</ref><box>[<257><163><487><357>]</box>"}, {"from": "human", "value": "Locate all regions matching: Books with visible titles: 'El Quijote' and 'La Comunidad' near the top."}, {"from": "gpt", "value": "<ref>Books with visible titles: 'El Quijote' and 'La Comunidad' near the top</ref><box>[<86><163><264><365>, <405><390><697><573>]</box>"}, {"from": "human", "value": "Find and box every: A colorful VHS tape labeled 'Video' with rainbow stripes next to the camera."}, {"from": "gpt", "value": "<ref>A colorful VHS tape labeled 'Video' with rainbow stripes next to the camera</ref><box>[<514><157><669><394>]</box>"}]}
+{"image": ["media/521470_497e939ec0c7d446.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw boxes around all: tall green cactus with spines on the left."}, {"from": "gpt", "value": "<ref>tall green cactus with spines on the left</ref><box>[<229><0><344><752>]</box>"}, {"from": "human", "value": "Detect all instances of: spotted snake-like objects in the bottom right foreground and return their bounding boxes."}, {"from": "gpt", "value": "<ref>spotted snake-like objects in the bottom right foreground</ref><box>[<609><928><782><997>, <748><685><902><997>, <887><642><961><997>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all instances of: the window in the background showing green trees outside."}, {"from": "gpt", "value": "<ref>the window in the background showing green trees outside</ref><box>[<366><0><490><492>, <556><0><697><385>]</box>"}]}
+{"image": ["media/523882_99a596bc73cd915d.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every occurrence of: A large yellow fish with orange-tipped fins swimming in an aquarium."}, {"from": "gpt", "value": "<ref>A large yellow fish with orange-tipped fins swimming in an aquarium</ref><box>[<16><133><768><727>]</box>"}, {"from": "human", "value": "Detect all instances of: The fish's body covered in hexagonal scales and return their bounding boxes."}, {"from": "gpt", "value": "<ref>The fish's body covered in hexagonal scales</ref><box>[<16><133><768><727>]</box>"}, {"from": "human", "value": "Draw boxes around all: The white label attached to a branch in the background."}, {"from": "gpt", "value": "<ref>The white label attached to a branch in the background</ref><box>[<738><308><824><463>]</box>"}]}
+{"image": ["media/584151_53a58ce859b6cce0.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn box coordinates for all: pink pig-shaped cake with blue dollar sign on its side."}, {"from": "gpt", "value": "<ref>pink pig-shaped cake with blue dollar sign on its side</ref><box>[<241><201><709><730>]</box>"}, {"from": "human", "value": "Return box coordinates for all: blue reusable water bottle next to a purple cup."}, {"from": "gpt", "value": "<ref>blue reusable water bottle next to a purple cup</ref><box>[<79><430><209><805>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: red water bottle behind the cake."}, {"from": "gpt", "value": "<ref>red water bottle behind the cake</ref><box>[<492><109><562><240>]</box>"}, {"from": "human", "value": "Detect all: wooden picnic table with cake and drinks on it."}, {"from": "gpt", "value": "<ref>wooden picnic table with cake and drinks on it</ref><box>[<0><459><998><997>]</box>"}]}
+{"image": ["media/711569_becaf09e5a791cbd.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every: UN TOUR AVEC LEGO GOLDMAN sign on the front of the model."}, {"from": "gpt", "value": "<ref>UN TOUR AVEC LEGO GOLDMAN sign on the front of the model</ref><box>[<446><594><673><651>]</box>"}, {"from": "human", "value": "Find and box every: the stage area with LEGO musicians and large speakers."}, {"from": "gpt", "value": "<ref>the stage area with LEGO musicians and large speakers</ref><box>[<404><157><673><429>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: crowd of LEGO minifigures facing the stage."}, {"from": "gpt", "value": "<ref>crowd of LEGO minifigures facing the stage</ref><box>[<209><415><938><593>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all: two white electric guitar silhouettes on the front panel."}, {"from": "gpt", "value": "<ref>two white electric guitar silhouettes on the front panel</ref><box>[<345><594><429><658>, <687><586><775><651>]</box>"}, {"from": "human", "value": "Outline all regions of: the poster to the left showing 'EN CONCERT 27, 28 et 29 juin 2014 à Rosheim'."}, {"from": "gpt", "value": "<ref>the poster to the left showing 'EN CONCERT 27, 28 et 29 juin 2014 à Rosheim'</ref><box>[<0><0><237><683>]</box>"}]}
+{"image": ["media/844358_54c04fa7f367a1c3.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all: Three headshots pinned vertically on the left wall."}, {"from": "gpt", "value": "<ref>Three headshots pinned vertically on the left wall</ref><box>[<108><312><310><536>, <286><69><459><329>, <319><320><519><581>]</box>"}, {"from": "human", "value": "Outline all regions of: The photo of a smiling person with glasses at the bottom center."}, {"from": "gpt", "value": "<ref>The photo of a smiling person with glasses at the bottom center</ref><box>[<341><598><509><839>]</box>"}]}
+{"image": ["media/884749_60f3c61253e0917a.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all regions matching: Cooked lobsters piled together in a white container."}, {"from": "gpt", "value": "<ref>Cooked lobsters piled together in a white container</ref><box>[<0><0><1000><1000>]</box>"}]}
+{"image": ["media/940810_44aa00c84350fdbd.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every: the beanbag with zebra print in the foreground."}, {"from": "gpt", "value": "<ref>the beanbag with zebra print in the foreground</ref><box>[<148><662><388><997>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: pink-haired person in blue shirt sitting on a floral beanbag."}, {"from": "gpt", "value": "<ref>pink-haired person in blue shirt sitting on a floral beanbag</ref><box>[<600><610><774><914>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all: winged figure wearing white tank top on a red-and-white beanbag."}, {"from": "gpt", "value": "<ref>winged figure wearing white tank top on a red-and-white beanbag</ref><box>[<741><248><826><453>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: large round floral rug in the center surrounded by beanbags."}, {"from": "gpt", "value": "<ref>large round floral rug in the center surrounded by beanbags</ref><box>[<337><386><782><740>]</box>"}]}
+{"image": ["media/970201_1c7f06f6d033ae0d.jpg"], "conversations": [{"from": "human", "value": "<image>\nProvide bounding boxes for every: A brown monkey sitting under a red metal bench."}, {"from": "gpt", "value": "<ref>A brown monkey sitting under a red metal bench</ref><box>[<46><515><632><921>]</box>"}, {"from": "human", "value": "Detect all: The red bench with a white smiley face on the backrest."}, {"from": "gpt", "value": "<ref>The red bench with a white smiley face on the backrest</ref><box>[<176><67><928><857>]</box>"}, {"from": "human", "value": "Locate all regions matching: Monkey's arm reaching under the front slat of the bench."}, {"from": "gpt", "value": "<ref>Monkey's arm reaching under the front slat of the bench</ref><box>[<337><583><633><784>]</box>"}, {"from": "human", "value": "Find all bounding boxes for: The gravel ground surrounding the bench and monkey."}, {"from": "gpt", "value": "<ref>The gravel ground surrounding the bench and monkey</ref><box>[<0><23><997><998>]</box>"}]}
+{"image": ["media/1124672_fe155561ca912833.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every occurrence of: person with a blue backpack standing on rocky terrain."}, {"from": "gpt", "value": "<ref>person with a blue backpack standing on rocky terrain</ref><box>[<505><685><540><845>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all instances of: rocky mountain ridge extending from the foreground to the center."}, {"from": "gpt", "value": "<ref>rocky mountain ridge extending from the foreground to the center</ref><box>[<0><333><934><997>]</box>"}, {"from": "human", "value": "Locate all regions matching: clouds below the mountain peak on the left."}, {"from": "gpt", "value": "<ref>clouds below the mountain peak on the left</ref><box>[<0><364><491><934>]</box>"}, {"from": "human", "value": "Detect all instances of: blue sky with thin white clouds above the ridge and return their bounding boxes."}, {"from": "gpt", "value": "<ref>blue sky with thin white clouds above the ridge</ref><box>[<0><2><998><395>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: a cluster of gray rocks in the foreground with yellow lichen."}, {"from": "gpt", "value": "<ref>a cluster of gray rocks in the foreground with yellow lichen</ref><box>[<149><430><998><997>]</box>"}]}
+{"image": ["media/1355137_33ad5f47f167ecfa.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all: the white tower with the clock and 'EXCELSIOR LIFE INSURANCE CO BLDG.' text."}, {"from": "gpt", "value": "<ref>the white tower with the clock and 'EXCELSIOR LIFE INSURANCE CO BLDG.' text</ref><box>[<8><273><168><816>]</box>"}, {"from": "human", "value": "Give bounding boxes for each: the red car parked near the corner of the building on the right."}, {"from": "gpt", "value": "<ref>the red car parked near the corner of the building on the right</ref><box>[<848><715><991><828>]</box>"}, {"from": "human", "value": "Return box coordinates for all: American flag on top of the tower."}, {"from": "gpt", "value": "<ref>American flag on top of the tower</ref><box>[<75><161><119><215>]</box>"}, {"from": "human", "value": "Locate all regions matching: the brick building with a flat roof and corner entrance."}, {"from": "gpt", "value": "<ref>the brick building with a flat roof and corner entrance</ref><box>[<149><180><979><834>]</box>"}]}
+{"image": ["media/1370395_62bec1f4f81aed7d.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw bounding boxes around all instances of: the white tower with a spherical structure on top on the left."}, {"from": "gpt", "value": "<ref>the white tower with a spherical structure on top on the left</ref><box>[<307><400><357><730>]</box>"}, {"from": "human", "value": "Outline all regions that contain: a cluster of lit buildings to the right of the central skyscraper."}, {"from": "gpt", "value": "<ref>a cluster of lit buildings to the right of the central skyscraper</ref><box>[<594><473><702><730>, <697><509><782><727>, <774><520><852><725>, <853><491><947><726>, <949><465><998><728>]</box>"}]}
+{"image": ["media/1412147_464f21333db8f564.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all: person in black graduation gown standing next to a woman in a beige coat."}, {"from": "gpt", "value": "<ref>person in black graduation gown standing next to a woman in a beige coat</ref><box>[<542><348><800><1000>]</box>"}, {"from": "human", "value": "Return box coordinates for all: small child in dark jacket sitting on the woman's hip."}, {"from": "gpt", "value": "<ref>small child in dark jacket sitting on the woman's hip</ref><box>[<347><472><510><928>]</box>"}, {"from": "human", "value": "Detect all: stone steps with metal handrails to the left of the group."}, {"from": "gpt", "value": "<ref>stone steps with metal handrails to the left of the group</ref><box>[<0><337><329><656>]</box>"}, {"from": "human", "value": "Draw boxes around all: large evergreen tree behind the group on the right."}, {"from": "gpt", "value": "<ref>large evergreen tree behind the group on the right</ref><box>[<407><0><998><626>]</box>"}]}

RoboAfford_pointing/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "RoboAfford_pointing",
+  "modality": [
+    "image"
+  ],
+  "category": "pointing",
+  "total": 62803,
+  "sampled": 20
+}

RoboAfford_pointing/sampled.jsonl

@@ -0,0 +1,20 @@
+{"image": ["media/002361_c39fbf87d8dddb617289c5b1e5d5e8c8412d3024c289b646593739b8de871f49.jpg"], "conversations": [{"from": "human", "value": "<image>\nPoint to the center of: the floor."}, {"from": "gpt", "value": "<ref>the floor</ref><box>[<259><458>]</box>"}, {"from": "human", "value": "Tap on: floor design."}, {"from": "gpt", "value": "<ref>floor design</ref><box>[<457><612>]</box>"}, {"from": "human", "value": "Point to the center of: columns."}, {"from": "gpt", "value": "<ref>columns</ref><box>[<700><561>, <743><542>]</box>"}], "width": 1200, "height": 1200}
+{"image": ["media/002741_3118defd138441c6c5150a1f6be0f110f16185fdbc960e02f5a60d5a520c0055.jpg"], "conversations": [{"from": "human", "value": "<image>\nPoint to: iphone charger cord."}, {"from": "gpt", "value": "<ref>iphone charger cord</ref><box>[<383><489>]</box>"}], "width": 756, "height": 1008}
+{"image": ["media/008352_2a80dd4bea3cc4470815c4a891dece7a4a74fc8c6bc06e7156eda26b027ff32d.jpg"], "conversations": [{"from": "human", "value": "<image>\nGive the point coordinate of: statue."}, {"from": "gpt", "value": "<ref>statue</ref><box>[<511><368>]</box>"}, {"from": "human", "value": "Mark a point on: plinth."}, {"from": "gpt", "value": "<ref>plinth</ref><box>[<588><647>]</box>"}], "width": 571, "height": 892}
+{"image": ["media/009705_c0b8a5ade2dbcf3be77c74336de22b88a5d301c542f45814cdb6ad97434747cb.jpg"], "conversations": [{"from": "human", "value": "<image>\nPoint to: chest."}, {"from": "gpt", "value": "<ref>chest</ref><box>[<551><276>]</box>"}, {"from": "human", "value": "Give the point coordinate of: light."}, {"from": "gpt", "value": "<ref>light</ref><box>[<14><766>, <263><751>, <433><724>, <751><705>, <909><689>]</box>"}, {"from": "human", "value": "Point to: backdrop."}, {"from": "gpt", "value": "<ref>backdrop</ref><box>[<741><392>]</box>"}], "width": 396, "height": 595}
+{"image": ["media/009768_20b7087e1b16384990ab5c186d1f6299360113ca91849a9e41c4ff834651b2a4.jpg"], "conversations": [{"from": "human", "value": "<image>\nClick on: corner."}, {"from": "gpt", "value": "<ref>corner</ref><box>[<76><274>, <84><711>, <972><186>, <981><776>]</box>"}], "width": 5312, "height": 2988}
+{"image": ["media/020202_0f69d554c05319521623f714d9c6434446248e0adca487aa2c279ee2fa8815ea.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutput a point location for: ceiling."}, {"from": "gpt", "value": "<ref>ceiling</ref><box>[<602><34>]</box>"}, {"from": "human", "value": "Point at the center of: glass door."}, {"from": "gpt", "value": "<ref>glass door</ref><box>[<211><393>, <230><390>, <318><389>, <344><392>, <364><392>]</box>"}, {"from": "human", "value": "Mark a point on: wall to the side of the door."}, {"from": "gpt", "value": "<ref>wall to the side of the door</ref><box>[<177><380>]</box>"}, {"from": "human", "value": "Give the point coordinate of: tile floor."}, {"from": "gpt", "value": "<ref>tile floor</ref><box>[<230><735>]</box>"}, {"from": "human", "value": "Point at the center of: shoeboxes on top shelf."}, {"from": "gpt", "value": "<ref>shoeboxes on top shelf</ref><box>[<880><345>, <885><332>, <905><326>, <905><345>, <906><312>, <935><312>, <935><324>, <935><338>, <935><345>]</box>"}], "width": 348, "height": 348}
+{"image": ["media/023897_f9020c20f12e20415ce8fceb942235ad163027cc7e7d75564d09b1b5d2211894.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutput a point for: wine bottle wrapper."}, {"from": "gpt", "value": "<ref>wine bottle wrapper</ref><box>[<514><89>]</box><ref>wine bottle wrapper</ref><box>[<516><375>]</box><ref>wine bottle wrapper</ref><box>[<530><153>]</box>"}, {"from": "human", "value": "Give the point coordinate of: green glass."}, {"from": "gpt", "value": "<ref>green glass</ref><box>[<557><652>]</box>"}, {"from": "human", "value": "Return a single point for: bottom edge of wine bottle wrapper."}, {"from": "gpt", "value": "<ref>bottom edge of wine bottle wrapper</ref><box>[<466><412>]</box>"}, {"from": "human", "value": "Click on: table."}, {"from": "gpt", "value": "<ref>table</ref><box>[<321><944>]</box>"}], "width": 799, "height": 1199}
+{"image": ["media/024196_7a615bdc39e3dee422f33165c10de817823eb3e673eb8411a5d03ea4ef336ad3.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutput a point for: horizontal line on wall."}, {"from": "gpt", "value": "<ref>horizontal line on wall</ref><box>[<26><68>]</box><ref>horizontal line on wall</ref><box>[<300><15>]</box><ref>horizontal line on wall</ref><box>[<901><35>]</box>"}, {"from": "human", "value": "Point to: top of shelf lip."}, {"from": "gpt", "value": "<ref>top of shelf lip</ref><box>[<487><808>]</box>"}, {"from": "human", "value": "Return a single point for: front of shelf lip."}, {"from": "gpt", "value": "<ref>front of shelf lip</ref><box>[<476><842>]</box>"}, {"from": "human", "value": "Point to the center of: pale blue shelf."}, {"from": "gpt", "value": "<ref>pale blue shelf</ref><box>[<618><157>]</box>"}, {"from": "human", "value": "Click on: tan flooring."}, {"from": "gpt", "value": "<ref>tan flooring</ref><box>[<52><938>]</box>"}], "width": 320, "height": 222}
+{"image": ["media/028166_35d5cbed6cee4be6960c1b0751174296e22bbfac05e5fc5110dae91bb07be37c.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutput a point for: door."}, {"from": "gpt", "value": "<ref>door</ref><box>[<756><584>]</box>"}, {"from": "human", "value": "Point to the center of: office."}, {"from": "gpt", "value": "<ref>office</ref><box>[<731><719>]</box>"}, {"from": "human", "value": "Tap on: picture."}, {"from": "gpt", "value": "<ref>picture</ref><box>[<741><768>]</box>"}, {"from": "human", "value": "Return a single point for: shower."}, {"from": "gpt", "value": "<ref>shower</ref><box>[<965><831>]</box>"}, {"from": "human", "value": "Tap on: wall."}, {"from": "gpt", "value": "<ref>wall</ref><box>[<512><634>]</box>"}], "width": 5376, "height": 2688}
+{"image": ["media/035755_0335bcefcb14ce1ef022c4ad39177a30e807d300ca749c909e88c1f8900a51f3.jpg"], "conversations": [{"from": "human", "value": "<image>\nGive the point coordinate of: wooden planter."}, {"from": "gpt", "value": "<ref>wooden planter</ref><box>[<867><704>]</box>"}, {"from": "human", "value": "Output a point location for: white pot."}, {"from": "gpt", "value": "<ref>white pot</ref><box>[<446><126>]</box>"}, {"from": "human", "value": "Click on: metal bench."}, {"from": "gpt", "value": "<ref>metal bench</ref><box>[<651><222>]</box>"}], "width": 1025, "height": 1366}
+{"image": ["media/036210_35b871cdc3f849f8c42e2de59e87dec3b8e5d7f9ef9278b71aef13c90eaba48f.jpg"], "conversations": [{"from": "human", "value": "<image>\nMark a point on: bottle."}, {"from": "gpt", "value": "<ref>bottle</ref><box>[<572><738>]</box>"}], "width": 500, "height": 500}
+{"image": ["media/038175_0a2e12fa16989c543f849d77223acb22ce0da15417455bbd98e46160cb1d3e04.jpg"], "conversations": [{"from": "human", "value": "<image>\nClick on: grey wall.\nExpected output: <ref>target name</ref><box>[[x, y]]</box>"}, {"from": "gpt", "value": "<ref>grey wall</ref><box>[<548><80>]</box>"}], "width": 1024, "height": 768}
+{"image": ["media/045286_c5c6fe560507abcdf0e603912680702d6b78e48e6fab20571eccc7b726a17498.jpg"], "conversations": [{"from": "human", "value": "<image>\nPoint at: dark pattern in wood."}, {"from": "gpt", "value": "<ref>dark pattern in wood</ref><box>[<970><869>]</box>"}, {"from": "human", "value": "Point at: wood cutting board."}, {"from": "gpt", "value": "<ref>wood cutting board</ref><box>[<907><920>]</box>"}, {"from": "human", "value": "Provide the center point of: table top."}, {"from": "gpt", "value": "<ref>table top</ref><box>[<945><413>]</box>"}], "width": 2048, "height": 1536}
+{"image": ["media/052174_e4eff2d5c60addf590742d0099d485157b1f72d7525734f95d5ddda550d9a776.jpg"], "conversations": [{"from": "human", "value": "<image>\nPoint to the center of: corners of wall art.\nUse format like: <ref>button</ref><box>[<350><200>]</box>"}, {"from": "gpt", "value": "<ref>corners of wall art</ref><box>[<65><267>, <70><85>, <209><81>, <213><264>]</box>"}, {"from": "human", "value": "Point to: speakers of small radio."}, {"from": "gpt", "value": "<ref>speakers of small radio</ref><box>[<566><735>, <620><736>]</box>"}, {"from": "human", "value": "Tap on: corners of side table surface."}, {"from": "gpt", "value": "<ref>corners of side table surface</ref><box>[<509><774>, <527><732>, <769><731>, <776><770>]</box>"}, {"from": "human", "value": "Point to: white background wall space."}, {"from": "gpt", "value": "<ref>white background wall space</ref><box>[<131><495>, <181><802>, <381><122>, <401><582>, <826><145>, <847><586>]</box>"}, {"from": "human", "value": "Output a point location for: corners of window panes."}, {"from": "gpt", "value": "<ref>corners of window panes</ref><box>[<919><225>, <924><145>, <933><513>]</box>"}], "width": 630, "height": 630}
+{"image": ["media/054696_30f1626154c42065b87e3f4e46ea07f3c8e62c15c83e45db3a5a69434879a052.jpg"], "conversations": [{"from": "human", "value": "<image>\nMark a point on: white arch."}, {"from": "gpt", "value": "<ref>white arch</ref><box>[<449><209>]</box><ref>white arch</ref><box>[<602><222>]</box>"}, {"from": "human", "value": "Click on: moulding."}, {"from": "gpt", "value": "<ref>moulding</ref><box>[<288><274>]</box><ref>moulding</ref><box>[<427><163>]</box><ref>moulding</ref><box>[<449><209>]</box><ref>moulding</ref><box>[<563><164>]</box><ref>moulding</ref><box>[<707><280>]</box>"}, {"from": "human", "value": "Provide the center point of: globe."}, {"from": "gpt", "value": "<ref>globe</ref><box>[<511><396>]</box>"}, {"from": "human", "value": "Point to: white pillars."}, {"from": "gpt", "value": "<ref>white pillars</ref><box>[<255><653>]</box><ref>white pillars</ref><box>[<739><670>]</box>"}, {"from": "human", "value": "Point to: door."}, {"from": "gpt", "value": "<ref>door</ref><box>[<496><706>]</box>"}], "width": 404, "height": 316}
+{"image": ["media/057875_701e616dbc284fadbe3d36c4993ee5a9eadd10bd921a62e2ec1bf541e769fbde.jpg"], "conversations": [{"from": "human", "value": "<image>\nPoint at: door."}, {"from": "gpt", "value": "<ref>door</ref><box>[<629><64>]</box>"}], "width": 1200, "height": 800}
+{"image": ["media/058915_6c6ec1045cd0992a83d8dff6bebf65e76b9e8f2a8027b9b2b9e03ad63b7b7395.jpg"], "conversations": [{"from": "human", "value": "<image>\nClick on: wires."}, {"from": "gpt", "value": "<ref>wires</ref><box>[<453><122>, <482><92>]</box>"}, {"from": "human", "value": "Tap on: floor joints."}, {"from": "gpt", "value": "<ref>floor joints</ref><box>[<405><74>, <454><138>, <479><297>]</box>"}, {"from": "human", "value": "Point at the center of: support pillar."}, {"from": "gpt", "value": "<ref>support pillar</ref><box>[<535><171>]</box>"}, {"from": "human", "value": "Give the point coordinate of: wall above metal rod."}, {"from": "gpt", "value": "<ref>wall above metal rod</ref><box>[<453><438>]</box>"}, {"from": "human", "value": "Point at the center of: light fixture."}, {"from": "gpt", "value": "<ref>light fixture</ref><box>[<595><47>]</box>"}], "width": 1024, "height": 576}
+{"image": ["media/059239_dbc60c7a790ad6dc85e4f5336443332f890e24a0e94cf3a9c742d4bc7f7fe26d.jpg"], "conversations": [{"from": "human", "value": "<image>\nTap on: lights."}, {"from": "gpt", "value": "<ref>lights</ref><box>[<235><452>]</box><ref>lights</ref><box>[<462><223>]</box><ref>lights</ref><box>[<878><423>]</box>"}], "width": 1280, "height": 720}
+{"image": ["media/061150_893c243bfbe94d0ae9972932a958fb4e46749a6c59711ad77f544b609ef15bac.jpg"], "conversations": [{"from": "human", "value": "<image>\nPoint to: gray upholstered chair."}, {"from": "gpt", "value": "<ref>gray upholstered chair</ref><box>[<443><754>]</box><ref>gray upholstered chair</ref><box>[<446><917>]</box><ref>gray upholstered chair</ref><box>[<506><698>]</box><ref>gray upholstered chair</ref><box>[<542><548>]</box><ref>gray upholstered chair</ref><box>[<542><608>]</box><ref>gray upholstered chair</ref><box>[<558><566>]</box><ref>gray upholstered chair</ref><box>[<577><545>]</box><ref>gray upholstered chair</ref><box>[<699><524>]</box>"}, {"from": "human", "value": "Point at: wooden chair."}, {"from": "gpt", "value": "<ref>wooden chair</ref><box>[<32><536>]</box><ref>wooden chair</ref><box>[<668><544>]</box><ref>wooden chair</ref><box>[<693><611>]</box><ref>wooden chair</ref><box>[<724><547>]</box><ref>wooden chair</ref><box>[<740><683>]</box><ref>wooden chair</ref><box>[<768><563>]</box><ref>wooden chair</ref><box>[<882><656>]</box><ref>wooden chair</ref><box>[<978><806>]</box><ref>wooden chair</ref><box>[<984><626>]</box>"}, {"from": "human", "value": "Give the point coordinate of: wooden napkin holder."}, {"from": "gpt", "value": "<ref>wooden napkin holder</ref><box>[<82><518>]</box><ref>wooden napkin holder</ref><box>[<308><683>]</box><ref>wooden napkin holder</ref><box>[<479><557>]</box><ref>wooden napkin holder</ref><box>[<523><521>]</box><ref>wooden napkin holder</ref><box>[<731><545>]</box><ref>wooden napkin holder</ref><box>[<829><592>]</box>"}, {"from": "human", "value": "Tap on: wooden table."}, {"from": "gpt", "value": "<ref>wooden table</ref><box>[<238><758>]</box><ref>wooden table</ref><box>[<461><580>]</box><ref>wooden table</ref><box>[<502><533>]</box><ref>wooden table</ref><box>[<668><553>]</box><ref>wooden table</ref><box>[<793><613>]</box>"}, {"from": "human", "value": "Return a single point for: sheer white curtain."}, {"from": "gpt", "value": "<ref>sheer white curtain</ref><box>[<85><305>]</box><ref>sheer white curtain</ref><box>[<272><326>]</box><ref>sheer white curtain</ref><box>[<367><314>]</box><ref>sheer white curtain</ref><box>[<550><320>]</box><ref>sheer white curtain</ref><box>[<666><320>]</box><ref>sheer white curtain</ref><box>[<769><305>]</box><ref>sheer white curtain</ref><box>[<941><215>]</box>"}], "width": 1000, "height": 667}
+{"image": ["media/061934_89c3f6f9c45eb3815f6b4a65fed3b69bab2b7b56ddcd7735910f8a645ee78ca9.jpg"], "conversations": [{"from": "human", "value": "<image>\nGive the point coordinate of: potted plant."}, {"from": "gpt", "value": "<ref>potted plant</ref><box>[<953><426>]</box><ref>potted plant</ref><box>[<978><424>]</box><ref>potted plant</ref><box>[<988><471>]</box>"}, {"from": "human", "value": "Point to: window."}, {"from": "gpt", "value": "<ref>window</ref><box>[<615><408>]</box><ref>window</ref><box>[<616><367>]</box><ref>window</ref><box>[<786><382>]</box><ref>window</ref><box>[<797><411>]</box><ref>window</ref><box>[<983><392>]</box>"}, {"from": "human", "value": "Point to: chimney."}, {"from": "gpt", "value": "<ref>chimney</ref><box>[<288><355>]</box><ref>chimney</ref><box>[<316><374>]</box>"}], "width": 426, "height": 640}

Roboafford_grounding/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "Roboafford_grounding",
+  "modality": [
+    "image"
+  ],
+  "category": "grounding",
+  "total": 64514,
+  "sampled": 20
+}

Roboafford_grounding/sampled.jsonl

@@ -0,0 +1,20 @@
+{"image": ["media/000232_000000296303.jpg"], "conversations": [{"from": "human", "value": "<image>\nGive the bounding box for each instance of: the object or appliance to hold and drink beverages."}, {"from": "gpt", "value": "<ref>the object or appliance to hold and drink beverages</ref><box>[<144><155><317><524>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all instances of: the object or appliance to drink from."}, {"from": "gpt", "value": "<ref>the object or appliance to drink from</ref><box>[<147><162><316><525>]</box>"}, {"from": "human", "value": "Outline all regions that contain: the object or appliance to sip beverages."}, {"from": "gpt", "value": "<ref>the object or appliance to sip beverages</ref><box>[<147><156><306><179>]</box>"}], "width": 640, "height": 480}
+{"image": ["media/000633_000000137451.jpg"], "conversations": [{"from": "human", "value": "<image>\nGive the bounding box for each instance of: the object or appliance to inspect the wheel area without causing damage."}, {"from": "gpt", "value": "<ref>the object or appliance to inspect the wheel area without causing damage</ref><box>[<88><542><155><608>, <439><572><452><608>, <889><564><998><622>]</box>"}, {"from": "human", "value": "Detect all instances of: the object or appliance to mount the tire securely and return their bounding boxes."}, {"from": "gpt", "value": "<ref>the object or appliance to mount the tire securely</ref><box>[<92><578><159><622>, <434><617><561><653>, <908><603><920><631>]</box>"}, {"from": "human", "value": "Find and box every occurrence of: the object or appliance to steer it in different directions."}, {"from": "gpt", "value": "<ref>the object or appliance to steer it in different directions</ref><box>[<91><572><217><625>, <436><606><567><650>, <903><594><998><639>]</box>"}, {"from": "human", "value": "Detect all instances of: the object or appliance to open the door and return their bounding boxes."}, {"from": "gpt", "value": "<ref>the object or appliance to open the door</ref><box>[<930><581><966><586>]</box>"}, {"from": "human", "value": "Find and box every occurrence of: the object or appliance to improve rear visibility while driving."}, {"from": "gpt", "value": "<ref>the object or appliance to improve rear visibility while driving</ref><box>[<125><550><139><561>, <972><550><983><569>]</box>"}], "width": 640, "height": 360}
+{"image": ["media/002889_000000029449.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind all bounding boxes for: the object or appliance to release soap."}, {"from": "gpt", "value": "<ref>the object or appliance to release soap</ref><box>[<923><673><956><708>]</box>"}, {"from": "human", "value": "Return box coordinates for all: the object or appliance to lather and cleanse the skin."}, {"from": "gpt", "value": "<ref>the object or appliance to lather and cleanse the skin</ref><box>[<904><703><948><759>]</box>"}], "width": 480, "height": 640}
+{"image": ["media/004931_000000578654.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutline all regions that contain: the object or appliance to illuminate a room."}, {"from": "gpt", "value": "<ref>the object or appliance to illuminate a room</ref><box>[<365><318><469><575>]</box>"}, {"from": "human", "value": "Outline all regions that contain: the object or appliance to control the direction of light."}, {"from": "gpt", "value": "<ref>the object or appliance to control the direction of light</ref><box>[<366><331><469><431>]</box>"}, {"from": "human", "value": "Detect all instances of: the object or appliance to secure the lampshade in place and return their bounding boxes."}, {"from": "gpt", "value": "<ref>the object or appliance to secure the lampshade in place</ref><box>[<408><319><417><331>]</box>"}, {"from": "human", "value": "Find all bounding boxes for: the object or appliance to adjust its height or position."}, {"from": "gpt", "value": "<ref>the object or appliance to adjust its height or position</ref><box>[<392><442><433><448>]</box>"}], "width": 640, "height": 480}
+{"image": ["media/007331_000000254240.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every: the object or appliance to clean up spills or wipe hands during a meal."}, {"from": "gpt", "value": "<ref>the object or appliance to clean up spills or wipe hands during a meal</ref><box>[<436><607><999><999>]</box><ref>the object or appliance to clean up spills or wipe hands during a meal</ref><box>[<465><313><654><418>]</box><ref>the object or appliance to clean up spills or wipe hands during a meal</ref><box>[<3><400><197><541>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all: the object or appliance to slice or cut various materials."}, {"from": "gpt", "value": "<ref>the object or appliance to slice or cut various materials</ref><box>[<817><753><938><997>]</box>"}, {"from": "human", "value": "Return box coordinates for all: the object or appliance to hold and drink beverages."}, {"from": "gpt", "value": "<ref>the object or appliance to hold and drink beverages</ref><box>[<801><176><962><723>]</box><ref>the object or appliance to hold and drink beverages</ref><box>[<606><168><735><524>]</box><ref>the object or appliance to hold and drink beverages</ref><box>[<231><85><333><472>]</box><ref>the object or appliance to hold and drink beverages</ref><box>[<131><162><256><492>]</box><ref>the object or appliance to hold and drink beverages</ref><box>[<210><164><258><343>]</box>"}, {"from": "human", "value": "Find all bounding boxes for: the object or appliance to cut safely."}, {"from": "gpt", "value": "<ref>the object or appliance to cut safely</ref><box>[<825><753><936><998>]</box>"}, {"from": "human", "value": "Return box coordinates for all: the object or appliance to hold liquid securely."}, {"from": "gpt", "value": "<ref>the object or appliance to hold liquid securely</ref><box>[<814><179><950><198>]</box>"}], "width": 640, "height": 425}
+{"image": ["media/008573_000000412584.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutline all regions of: the object or appliance to dispose of waste materials."}, {"from": "gpt", "value": "<ref>the object or appliance to dispose of waste materials</ref><box>[<30><498><49><536>]</box><ref>the object or appliance to dispose of waste materials</ref><box>[<317><516><341><553>]</box><ref>the object or appliance to dispose of waste materials</ref><box>[<17><497><37><535>]</box><ref>the object or appliance to dispose of waste materials</ref><box>[<46><498><69><537>]</box>"}, {"from": "human", "value": "Outline all regions of: the object or appliance to throw in garbage."}, {"from": "gpt", "value": "<ref>the object or appliance to throw in garbage</ref><box>[<30><496><47><501>]</box><ref>the object or appliance to throw in garbage</ref><box>[<317><518><339><525>]</box><ref>the object or appliance to throw in garbage</ref><box>[<17><496><36><501>]</box><ref>the object or appliance to throw in garbage</ref><box>[<47><499><69><501>]</box>"}], "width": 640, "height": 425}
+{"image": ["media/010589_000000244925.jpg"], "conversations": [{"from": "human", "value": "Provide bounding boxes for every: the object or appliance to support sitting comfortably."}, {"from": "gpt", "value": "<ref>the object or appliance to support sitting comfortably</ref><box>[<795><515><938><585>]</box><ref>the object or appliance to support sitting comfortably</ref><box>[<620><532><820><569>]</box>"}, {"from": "human", "value": "Return box coordinates for all: the object or appliance to stabilize its structure or provide additional support."}, {"from": "gpt", "value": "<ref>the object or appliance to stabilize its structure or provide additional support</ref><box>[<772><546><820><571>]</box>"}, {"from": "human", "value": "Draw boxes around all: the object or appliance to support it on the ground."}, {"from": "gpt", "value": "<ref>the object or appliance to support it on the ground</ref><box>[<772><504><898><625>]</box>"}], "width": 640, "height": 427}
+{"image": ["media/017167_000000137595.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all regions matching: the object or appliance to access its contents."}, {"from": "gpt", "value": "<ref>the object or appliance to access its contents</ref><box>[<498><658><555><704>]</box><ref>the object or appliance to access its contents</ref><box>[<552><690><691><760>]</box>"}], "width": 640, "height": 480}
+{"image": ["media/018132_000000477500.jpg"], "conversations": [{"from": "human", "value": "<image>\nProvide bounding boxes for every: the object or appliance to absorb impact during a collision."}, {"from": "gpt", "value": "<ref>the object or appliance to absorb impact during a collision</ref><box>[<896><502><926><556>]</box>"}, {"from": "human", "value": "Return box coordinates for all: the object or appliance to mount the tire securely."}, {"from": "gpt", "value": "<ref>the object or appliance to mount the tire securely</ref><box>[<932><517><960><574>]</box>"}, {"from": "human", "value": "Detect all: the object or appliance to store or hold items securely."}, {"from": "gpt", "value": "<ref>the object or appliance to store or hold items securely</ref><box>[<698><577><768><607>]</box>"}, {"from": "human", "value": "Give bounding boxes for each: the object or appliance to provide structural support when carrying items."}, {"from": "gpt", "value": "<ref>the object or appliance to provide structural support when carrying items</ref><box>[<696><583><806><676>]</box>"}], "width": 500, "height": 333}
+{"image": ["media/021859_000000518481.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn box coordinates for all: the object or appliance to support your feet while sitting."}, {"from": "gpt", "value": "<ref>the object or appliance to support your feet while sitting</ref><box>[<89><717><219><766>]</box>"}, {"from": "human", "value": "Return box coordinates for all: the object or appliance to control the direction of light."}, {"from": "gpt", "value": "<ref>the object or appliance to control the direction of light</ref><box>[<570><365><583><389>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all: the object or appliance to adjust its height or position."}, {"from": "gpt", "value": "<ref>the object or appliance to adjust its height or position</ref><box>[<572><384><575><431>]</box>"}], "width": 640, "height": 427}
+{"image": ["media/028028_000000064941.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind all bounding boxes for: the object or appliance to prevent injury."}, {"from": "gpt", "value": "<ref>the object or appliance to prevent injury</ref><box>[<183><319><258><354>, <561><273><640><319>, <722><410><845><499>]</box>"}, {"from": "human", "value": "Return bounding box coordinates for all: the object or appliance to the head."}, {"from": "gpt", "value": "<ref>the object or appliance to the head</ref><box>[<215><352><223><355>]</box>"}, {"from": "human", "value": "Detect all instances of: the object or appliance to enable movement and return their bounding boxes."}, {"from": "gpt", "value": "<ref>the object or appliance to enable movement</ref><box>[<200><548><262><764>, <656><670><894><964>]</box>"}, {"from": "human", "value": "Detect all instances of: the object or appliance to propel it forward and return their bounding boxes."}, {"from": "gpt", "value": "<ref>the object or appliance to propel it forward</ref><box>[<181><648><262><670>, <721><728><790><856>]</box>"}], "width": 480, "height": 640}
+{"image": ["media/028956_000000489100.jpg"], "conversations": [{"from": "human", "value": "<image>\nReturn bounding box coordinates for all: the object or appliance to provide back support."}, {"from": "gpt", "value": "<ref>the object or appliance to provide back support</ref><box>[<806><548><972><644>]</box>"}], "width": 423, "height": 564}
+{"image": ["media/036154_000000294100.jpg"], "conversations": [{"from": "human", "value": "<image>\nDetect all instances of: the object or appliance to store and transport various items securely and return their bounding boxes.\nUse format like: <ref>car</ref><box>[<120><50><400><300>]</box><ref>person</ref><box>[<500><100><650><450>]</box>"}, {"from": "gpt", "value": "<ref>the object or appliance to store and transport various items securely</ref><box>[<145><22><182><40>, <628><81><707><122>]</box>"}, {"from": "human", "value": "Find and box every occurrence of: the object or appliance to perform tasks and provide companionship."}, {"from": "gpt", "value": "<ref>the object or appliance to perform tasks and provide companionship</ref><box>[<69><206><778><459>]</box>"}, {"from": "human", "value": "Outline all regions that contain: the object or appliance to move it horizontally."}, {"from": "gpt", "value": "<ref>the object or appliance to move it horizontally</ref><box>[<146><23><182><39>, <628><81><705><122>]</box>"}, {"from": "human", "value": "Find all bounding boxes for: the object or appliance to lift and transport it easily."}, {"from": "gpt", "value": "<ref>the object or appliance to lift and transport it easily</ref><box>[<655><95><699><105>]</box>"}, {"from": "human", "value": "Find and box every occurrence of: the object or appliance to communicate with it."}, {"from": "gpt", "value": "<ref>the object or appliance to communicate with it</ref><box>[<71><309><188><380>]</box>"}], "width": 478, "height": 640}
+{"image": ["media/036759_000000396400.jpg"], "conversations": [{"from": "human", "value": "Draw boxes around all: the object or appliance to hear sounds.\nFormat example: <ref>dog</ref><box>[<200><150><500><400>]</box>"}, {"from": "gpt", "value": "<ref>the object or appliance to hear sounds</ref><box>[<542><429><828><902>]</box>"}], "width": 500, "height": 357}
+{"image": ["media/041753_000000363933.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutline all regions that contain: the object or appliance to change channels or adjust the volume on a television."}, {"from": "gpt", "value": "<ref>the object or appliance to change channels or adjust the volume on a television</ref><box>[<842><684><1000><761>]</box>"}, {"from": "human", "value": "Locate all regions matching: the object or appliance to operate its functions."}, {"from": "gpt", "value": "<ref>the object or appliance to operate its functions</ref><box>[<895><683><998><712>]</box>"}], "width": 640, "height": 480}
+{"image": ["media/046246_000000394640.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw boxes around all: the object or appliance to transport people from one location to another."}, {"from": "gpt", "value": "<ref>the object or appliance to transport people from one location to another</ref><box>[<148><239><239><310>, <414><279><452><316>, <444><283><484><316>]</box>"}, {"from": "human", "value": "Give the bounding box for each instance of: the object or appliance to maintain clear visibility while driving."}, {"from": "gpt", "value": "<ref>the object or appliance to maintain clear visibility while driving</ref><box>[<452><285><469><295>]</box>"}, {"from": "human", "value": "Locate all regions matching: the object or appliance to absorb impact during a collision."}, {"from": "gpt", "value": "<ref>the object or appliance to absorb impact during a collision</ref><box>[<444><295><466><313>]</box>"}, {"from": "human", "value": "Locate all regions matching: the object or appliance to inspect the wheel area without causing damage."}, {"from": "gpt", "value": "<ref>the object or appliance to inspect the wheel area without causing damage</ref><box>[<150><277><159><293>, <462><295><483><313>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all instances of: the object or appliance to allow fresh air to enter."}, {"from": "gpt", "value": "<ref>the object or appliance to allow fresh air to enter</ref><box>[<164><259><184><275>, <469><288><480><295>]</box>"}], "width": 640, "height": 386}
+{"image": ["media/048577_000000325821.jpg"], "conversations": [{"from": "human", "value": "<image>\nOutline all regions that contain: the object or appliance to open it or pour out its contents."}, {"from": "gpt", "value": "<ref>the object or appliance to open it or pour out its contents</ref><box>[<458><591><558><663>, <733><761><847><848>, <797><853><916><953>, <859><975><958><998>]</box>"}, {"from": "human", "value": "Provide bounding boxes for every: the object or appliance to hold it securely."}, {"from": "gpt", "value": "<ref>the object or appliance to hold it securely</ref><box>[<447><636><569><978>, <722><815><858><1000>, <781><913><925><998>]</box>"}, {"from": "human", "value": "Draw boxes around all: the object or appliance to store its contents securely."}, {"from": "gpt", "value": "<ref>the object or appliance to store its contents securely</ref><box>[<867><983><955><998>]</box>"}, {"from": "human", "value": "Detect all instances of: the object or appliance to place items on and return their bounding boxes."}, {"from": "gpt", "value": "<ref>the object or appliance to place items on</ref><box>[<58><603><586><995>]</box>"}, {"from": "human", "value": "Draw boxes around all: the object or appliance to reinforce its structure and provide support for the tabletop."}, {"from": "gpt", "value": "<ref>the object or appliance to reinforce its structure and provide support for the tabletop</ref><box>[<0><828><266><998>]</box>"}], "width": 640, "height": 401}
+{"image": ["media/050536_000000534468.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind and box every occurrence of: the object or appliance to view programs and shows."}, {"from": "gpt", "value": "<ref>the object or appliance to view programs and shows</ref><box>[<607><281><752><384>]</box>"}, {"from": "human", "value": "Give the bounding box for each instance of: the object or appliance to change channels or adjust the volume from a distance."}, {"from": "gpt", "value": "<ref>the object or appliance to change channels or adjust the volume from a distance</ref><box>[<390><441><421><463>, <447><558><491><604>]</box>"}, {"from": "human", "value": "Locate all regions matching: the object or appliance to write or draw on paper by applying pressure."}, {"from": "gpt", "value": "<ref>the object or appliance to write or draw on paper by applying pressure</ref><box>[<469><683><476><715>, <610><812><687><863>, <662><844><711><966>, <678><873><789><953>, <684><860><792><939>, <688><935><738><967>, <708><808><745><869>]</box>"}, {"from": "human", "value": "Locate all regions matching: the object or appliance to ensure stability."}, {"from": "gpt", "value": "<ref>the object or appliance to ensure stability</ref><box>[<608><371><748><383>]</box>"}, {"from": "human", "value": "Find and box every occurrence of: the object or appliance to replace the batteries."}, {"from": "gpt", "value": "<ref>the object or appliance to replace the batteries</ref><box>[<389><442><419><462>]</box>"}], "width": 640, "height": 480}
+{"image": ["media/054352_000000043813.jpg"], "conversations": [{"from": "human", "value": "<image>\nLocate all regions matching: the object or appliance to provide protection during activities like biking or construction.\nUse format like: <ref>car</ref><box>[<120><50><400><300>]</box><ref>person</ref><box>[<500><100><650><450>]</box>"}, {"from": "gpt", "value": "<ref>the object or appliance to provide protection during activities like biking or construction</ref><box>[<680><340><777><458>]</box>"}, {"from": "human", "value": "Find and box every occurrence of: the object or appliance to secure it on the head."}, {"from": "gpt", "value": "<ref>the object or appliance to secure it on the head</ref><box>[<688><381><777><455>]</box>"}, {"from": "human", "value": "Draw bounding boxes around all instances of: the object or appliance to ensure a snug fit."}, {"from": "gpt", "value": "<ref>the object or appliance to ensure a snug fit</ref><box>[<681><375><777><386>]</box>"}], "width": 480, "height": 640}
+{"image": ["media/064212_000000139214.jpg"], "conversations": [{"from": "human", "value": "<image>\nFind all bounding boxes for: the object or appliance to identify its brand or manufacturer visually."}, {"from": "gpt", "value": "<ref>the object or appliance to identify its brand or manufacturer visually</ref><box>[<755><886><783><913>]</box>"}], "width": 640, "height": 481}

ShareRobot_affordance/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "ShareRobot_affordance",
+  "modality": [
+    "image"
+  ],
+  "category": "affordance",
+  "total": 6522,
+  "sampled": 20
+}

ShareRobot_affordance/sampled.jsonl

@@ -0,0 +1,20 @@
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"pick up the red object from the table\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<472><632><565><689>]"}], "image": ["media/000162_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"take the blue fork and place it on the orange cloth\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<121><436><183><514>]"}], "image": ["media/000788_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"slide the checkered can so that it's closer to the silver pan, on the empty corner of the table.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<719><727><737><756>]"}], "image": ["media/000993_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"place eggplant to the right of the chicken wing\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<317><273><394><366>]"}], "image": ["media/001629_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"place broccoli in a pot\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<95><488><129><541>]"}], "image": ["media/002111_frame_15.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"move the pot to the bottom right of the table\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<374><323><415><402>]"}], "image": ["media/002128_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"put the cucumber in the red pot.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<178><436><221><479>]"}], "image": ["media/002741_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"move the bowl to the top left corner of table\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<492><596><558><707>]"}], "image": ["media/002819_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"place the metal pot on the yellow cloth.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<388><423><464><525>]"}], "image": ["media/002876_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"put the brush to the right of the sushi.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<173><414><236><488>]"}], "image": ["media/002951_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"move yellow cloth closer to the left of silver pot.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<133><400><248><511>]"}], "image": ["media/002957_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"pick the white bottle and place it.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<374><173><526><291>]"}], "image": ["media/003373_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"close the door.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<445><579><488><623>]"}], "image": ["media/003685_frame_15.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"close the door.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<564><467><661><571>]"}], "image": ["media/003903_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"pick ketchup from the table and place in oven.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<571><508><631><584>]"}], "image": ["media/004566_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"drag mug backwards.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<618><136><682><212>]"}], "image": ["media/004590_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"drag mug right to left.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<583><621><608><654>]"}], "image": ["media/004960_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"pick ketchup from strainer and place it on the table\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<391><614><442><682>]"}], "image": ["media/005159_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"pick banana from table and place on oven.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<558><536><621><651>]"}], "image": ["media/005567_frame_0.png"]}
+{"conversations": [{"from": "human", "value": "<image>\nYou are a robot. The task is \"pick banana and place it in mug.\". Please predict a possible affordance area of the end effector? Output the bounding box as <x1><y1><x2><y2> with coordinates normalized to [0, 1000], where (x1,y1) is the top-left and (x2,y2) is the bottom-right corner."}, {"from": "gpt", "value": "[<567><530><631><640>]"}], "image": ["media/006099_frame_0.png"]}

agibot_traj_image_500000_trajkmeans_train/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "agibot_traj_image_500000_trajkmeans_train",
+  "modality": [
+    "image"
+  ],
+  "category": "embodied_trajectory",
+  "total": 496837,
+  "sampled": 20
+}

codalm_train_converted_en/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "codalm_train_converted_en",
+  "modality": [
+    "image"
+  ],
+  "category": "drive_qa",
+  "total": 20495,
+  "sampled": 20
+}

codalm_train_converted_zh/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "codalm_train_converted_zh",
+  "modality": [
+    "image"
+  ],
+  "category": "drive_qa",
+  "total": 20495,
+  "sampled": 20
+}

codalm_train_converted_zh/sampled.jsonl

@@ -0,0 +1,20 @@
+{"id": "Train_general_292", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请描述每个物体的外观、位置、方向，并解释为什么它会影响自车的行为。"}, {"from": "gpt", "value": "在JSON数据描述的交通图像中，有几个关键的道路使用者和物体需要注意：\n\n1. **车辆**：\n   - 最左侧车道上有一辆白色汽车，似乎正在前进或准备这样做。这影响了自车（ego vehicle）的决策，特别是如果它打算使用左侧车道。\n   - 几辆车在交叉路口的不同车道上，其中一些面向自车。这表明了多车道和交通流向，需要自车做好准备，以便在交叉路口穿越或通过复杂交通。\n\n2. **脆弱道路使用者**：\n   - 在道路右侧可以看到一群行人和骑自行车的人，其中一些人可能准备过马路。自车需要密切监视这些道路使用者，以防他们进入人行横道，因为他们有可能穿过自车前方。\n\n3. **交通标志**：\n   - 在图像的右侧，交通灯杆上装有一个禁止左转的标志。这通知自车和其他驾驶员在这个交叉路口禁止左转。\n\n4. **交通灯**：\n   - 现场可见的交通灯对直行亮绿灯，但左转时显示红灯箭头。这告知自车可以直行前进，但不能左转。\n\n5. **其他物体**：\n   - 在交叉路口左侧远处可以看到一台建筑起重机。虽然它不直接影响即时驾驶决策，但它标志着附近的施工活动，暗示可能有施工区或交通模式变化。"}], "image": ["media/000292_Train_general_292.jpg"]}
+{"id": "Train_general_493", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请描述每个物体的外观、位置、方向，并解释为什么它会影响自车的行为。"}, {"from": "gpt", "value": "在观察到的交通图像中，有几个显著的道路使用者和元素。首先，两名行人在自车的左侧，紧邻一个有栏杆的门，背对自车行走。尽管这些行人目前不在道路上，但他们足够接近，可能会意外地走进行车路径，影响自车的行为。\n\n此外，在自车的右侧有一个蓝色圆形标志，中间有一条红色斜线穿过摩托车的符号。这个交通标志表明摩托车禁止进入这个区域，告知自车不应预期从那个方向来的摩托车。\n\n自车右侧还放置了橙色交通锥，一直延伸到有栏杆的门区域。这些锥桶很可能标记了一个受限区域，或者表示某种形式的道路施工或障碍，提示自车保持安全距离，避免进入该区域。\n\n此外，一个障碍门已经放下，阻挡了自车前方左侧的道路通行。这个障碍物实际上阻止了车辆进入，要求自车停下或寻找另一条路线，如果这是预定的路径的话。"}], "image": ["media/000493_Train_general_493.jpg"]}
+{"id": "Train_general_1376", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请描述每个物体的外观、位置、方向，并解释为什么它会影响自车的行为。"}, {"from": "gpt", "value": "在交通图像中，没有车辆出现。然而，存在易受伤害的道路使用者；具体来说，行人正在过马路，其中一人位于直行的自车道路中央。这种情况要求车辆必须停下或减速以确保行人安全通过。\n\n虽然图像中看不到交通标志，但前方有红绿灯。由于图像质量和距离的原因，红绿灯的当前状态不明确，但自驾车必须准备好适当地响应灯号，特别是考虑到行人的存在。\n\n道路左侧设置了交通锥，围成一个障碍，周围似乎是一个检查站或路边活动区。这种布置表明可能有一部分道路受到限制，车辆应该��备好可能的改道或道路空间减少。\n\n图像中没有出现障碍物。然而，还有其他一些值得关注的物体，包括一个帐篷和一个小结构物，位于道路旁的行人过街区附近。这些可能是出于安全或信息目的。虽然它们不会直接影响驾驶行为，但这些结构物的存在可能会影响行人的活动，因此，车辆应该监控可能接近或离开这一区域的个体。"}], "image": ["media/001376_Train_general_1376.jpg"]}
+{"id": "Train_general_2932", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请描述每个物体的外观、位置、方向，并解释为什么它会影响自车的行为。"}, {"from": "gpt", "value": "根据JSON数据描述的图像显示，有一辆占据右侧车道的大型卡车显著出现，该车辆带有明显的危险警示标记以及红白色的后下护栏系统。这辆车表明前方可能有施工区域，它的体积可能会遮挡道路标志和其他道路元素的视线，因此需要驾驶员提高警惕和意识，以影响他们的驾驶行为。\n\n另外，道路一侧放置了红白条纹隔离栏，表明右侧车道已关闭或正在进行道路施工。因此，包括本车在内的驾驶员应做好准备，向左并线以避开这一障碍，确保他们的旅程能够安全继续。\n\n图像中还可以看到右侧有便携式隔离栏，但从图像的角度只能看到部分。它们的存在表明交通的正常流动可能会临时改变。这些隔离栏不仅指引车辆行驶，也暗示附近可能有工人或设备可能会意外侵入道路。\n\n此外，图像中有一个亮光源的存在，很可能来自路灯，它引入了眩光因素，可能会通过短暂地使驾驶员目眩或阻碍道路使用者和交通信号的能见度，影响驾驶条件。\n\n在这张特定的交通图像数据中，并未显示有易受危险的道路使用者、交通灯或交通锥标识。"}], "image": ["media/002932_Train_general_2932.jpg"]}
+{"id": "Train_general_3700", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请描述每个物体的外观、位置、方向，并解释为什么它会影响自车的行为。"}, {"from": "gpt", "value": "在观察到的交通场景中，显著地注意到一辆大型车辆的存在，很可能是一辆卡车，它位于与本车同一车道上，并且与本车同向行驶。关于这辆大型车辆需要考虑的关键因素是它的尺寸以及与本车的近距离。这就要求本车保持安全的跟车距离，以应对前方卡车可能突然进行的任何停车或机动动作。场景中缺乏脆弱路面使用者、交通标志、交通灯、交通锥、障碍物以及其他物体的存在，表明这是一个主要由本车与前方卡车之间的车辆动态所主导的场景。"}], "image": ["media/003700_Train_general_3700.jpg"]}
+{"id": "Train_general_3920", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请描述每个物体的外观、位置、方向，并解释为什么它会影响自车的行为。"}, {"from": "gpt", "value": "在观察到的交通图像中，未见任何车辆、易受伤害的道路使用者、交通标志、交通信号灯、障碍物或其他物体存在。然而，值得注意的是有一排交通锥。这些锥标是排列在道路右侧的一系列橙色交通锥，用来标记关闭的一段车道。锥标的设置清楚地表明，道路的一部分目前不可通行，这可能是由于正在进行的道路施工或现有的危险状况。因此，这种情况要求自动驾驶车辆，亦即自驾车，必须调整其行驶路径，以避开关闭的车道部分，安全地继续行驶。"}], "image": ["media/003920_Train_general_3920.jpg"]}
+{"id": "Train_regional_194", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "一个圆形的直立交通管理物体，可能用于创建视觉障碍或指导交通流向。自车应避开这些物体，以防损坏车辆或物体，并遵循可能的交通模式变化。"}], "image": ["media/005078_Train_regional_194.jpg"]}
+{"id": "Train_regional_600", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这是一辆公交车，是一种为沿固定路线载运众多乘客而设计的大型机动车辆。公交车的存在会影响私家车司机的驾驶行为，因为他们需要遵守与大型车辆相关的交通规则，例如保持安全的车距、注意更宽的转弯半径，以及如果公交车正在服务中可能需要准备停车等待更长时间。"}], "image": ["media/005484_Train_regional_600.jpg"]}
+{"id": "Train_regional_2848", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这个对象是一种屏障，可能由塑料或其它轻质材料制成，用以标明建筑区域或指引车辆和行人绕过危险区域。类似的屏障对于确保道路使用者不会无意中进入危险区域至关重要，从而保护他们和现场工人的安全。司机应当小心行驶，注意观察这些屏障，不要试图绕过它们。"}], "image": ["media/007732_Train_regional_2848.jpg"]}
+{"id": "Train_regional_3081", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这个物体是一辆停在路边的摩托车。摩托车是小型机动车辆，其出现可能表明此区域人口稠密或繁忙。自驾车应保持安全距离，因为摩托车可能会突然启动。另外，车辆还应注意周围的其他类似小型车辆和行人。"}], "image": ["media/007965_Train_regional_3081.jpg"]}
+{"id": "Train_regional_3266", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这个物体是一种建筑车辆，特别是一台装载机。它的显著特征包括黄色外观和一个大型前斗，用于移动大量土壤或物料。在驾驶场景中，像这样的建筑车辆通常意味着附近有建筑活动，可能导致道路布局变化、临时交通信号或改道。驾驶员需要保持警惕，减速行驶，并准备应对意外障碍或道路条件的变化。"}], "image": ["media/008150_Train_regional_3266.jpg"]}
+{"id": "Train_regional_4163", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这是一辆工程车辆，具体是一辆混凝土搅拌车，其显著特征是一个大型旋转鼓，用于搅拌混凝土。这种车辆的出现表明该区域有建筑工程，可能会导致交通减慢、潜在的绕行或短暂停车。包括自车在内的驾驶员应对工程车辆的突然移动或附近的建筑工人和材料保持警惕。"}], "image": ["media/009047_Train_regional_4163.jpg"]}
+{"id": "Train_regional_5749", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这个物体是位于路边的垃圾桶。虽然它并不直接阻碍车辆行驶，但由于其靠近路缘的位置，要求司机在经过时特别注意侧向间隙，尤其是在车道较窄或绕行其他道路使用者如自行车手时。"}], "image": ["media/010633_Train_regional_5749.jpg"]}
+{"id": "Train_regional_7285", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这张图片展示了一辆机动踏板摩托车，它是一种两轮机动车。摩托车通常行驶速度比汽车慢，且能迅速机动。自车需要注意摩托车的存在，可能需要减速并保持警惕，以防摩托车在交通或拥挤的城市环境中突然变动或转弯。"}], "image": ["media/012169_Train_regional_7285.jpg"]}
+{"id": "Train_regional_8085", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "一辆装载着看似货物的三轮车停在路边。对于本车来说，这可能意味着需要小心，因为三轮车可能会突然并入交通流或进行装卸等活动。这强调了本车在经过三轮车附近时保持安全距离并可能减速的重要性。"}], "image": ["media/012969_Train_regional_8085.jpg"]}
+{"id": "Train_regional_9342", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这个物体是一辆公交车，以其大尺寸和矩形形状为特征，通常用于沿指定路线运输众多乘客。公交车在路上的出现表明它可能会频繁停车以接送乘客，影响交通流。驾驶员应保持安全的跟车距离，并在接近或驾驶经过公交车时准备好可能的突然停车或减速。"}], "image": ["media/014226_Train_regional_9342.jpg"]}
+{"id": "Train_regional_10689", "conversations": [{"from": "human", "value": "<image>\n请描述图像中红色矩形内的物体，并解释为什么它会影响自车的驾驶。"}, {"from": "gpt", "value": "这个物体看起来是一堆杂项材料，可能与建筑相关，放置在路边。这样的物料出现在路上或路边可能意味着正在进行的建设活动或临时的道路状况，可能需要驾驶员加倍小心，减速或暂时遵循改变的交通模式。"}], "image": ["media/015573_Train_regional_10689.jpg"]}
+{"id": "Train_suggestion_3429", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请根据当前场景为自车提供驾驶建议。"}, {"from": "gpt", "value": "保持当前车道并在必要时减速，以便因前方左侧车道关闭而可能进行的合并或车道变换做好准备。"}], "image": ["media/019040_Train_suggestion_3429.jpg"]}
+{"id": "Train_suggestion_3548", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请根据当前场景为自车提供驾驶建议。"}, {"from": "gpt", "value": "自车应保持安全的跟车距离，尤其是与平板卡车保持距离，以便有时间对任何掉落的碎片做出反应。此外，如果水泥搅拌车开始并线，要准备减速或者变道。建议定期检查后视镜和盲点，以保持对周围环境的认知。"}], "image": ["media/019159_Train_suggestion_3548.jpg"]}
+{"id": "Train_suggestion_3925", "conversations": [{"from": "human", "value": "<image>\n这是一张从自车视角捕捉的交通图像。重点关注影响自车驾驶行为的物体：车辆（轿车、卡车、公交车等）、易受伤害的道路使用者（行人、自行车骑行者、摩托车骑行者）、交通标志（禁止停车、警告、指示等）、交通信号灯（红灯、绿灯、黄灯）、交通锥、障碍物，以及杂物（碎片、垃圾桶、动物等）。请不要讨论以上七类以外的任何物体。请根据当前场景为自车提供驾驶建议。"}, {"from": "gpt", "value": "由于前方红灯，本车应减速并准备停车。同时，由于水泥搅拌车体积庞大且占据多个车道，保持安全的跟车距离至关重要。本车应密切监控搅拌车的动态，以便对任何车道变换或突然动作做出适当反应。"}], "image": ["media/019536_Train_suggestion_3925.jpg"]}

egoit_ADL/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "egoit_ADL",
+  "modality": [
+    "video"
+  ],
+  "category": "embodied_video_vqa",
+  "total": 3004,
+  "sampled": 20
+}

egoit_ADL/sampled.jsonl

@@ -0,0 +1,20 @@
+{"id": "P_13_long_22_0", "conversations": [{"from": "human", "value": "<video>\nCan you summarize the overall intent of the activities described in the video, and how do they collectively contribute to the main goal of the individual?\nA. The individual aims to host a social gathering, preparing snacks and drinks for the guests.\nB. The main intent is to create a detailed schedule for the upcoming week, aiming to boost productivity.\nC. The primary purpose is to relax and unwind.\nD. The goal is to carry out a deep cleaning of the entire house, focusing on maintaining a spotless environment.\nE. The activities aim to prepare for an intense workout session, aligning with a fitness-focused goal.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "C"}, {"from": "human", "value": "How do the described actions reflect the individual's approach to maintaining their environment, and what does this suggest about their priorities?\nA. The actions show a meticulous approach to organizing every item in the house, suggesting an obsession with order.\nB. The actions illustrate a balance between relaxation and cleanliness.\nC. The behaviors indicate a chaotic lifestyle, with random actions taken without any clear intent or purpose.\nD. The activities demonstrate a complete disregard for cleanliness, focusing solely on personal entertainment and leisure.\nE. The actions reveal a tendency to multitask excessively, prioritizing efficiency over relaxation.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "B"}, {"from": "human", "value": "Identify the primary activity in the video and explain its significance in relation to the individual's overall routine.\nA. Organizing household items is the primary activity, highlighting a priority on tidiness and order.\nB. Preparing a meal is the main activity, showing an emphasis on culinary skills and nutrition.\nC. Engaging in a hobby like painting or crafting is the main activity, indicating a creative outlet.\nD. Cleaning the kitchen is the primary activity, reflecting a focus on maintaining a pristine environment.\nE. Watching TV while eating and drinking is the main activity, signifying a relaxing break.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "E"}, {"from": "human", "value": "Compare the activities taking place in the video and explain how they collectively create a sense of the individual's lifestyle and habits.\nA. The actions suggest a lifestyle centered around social interactions and frequent gatherings.\nB. The activities reveal a high-stress lifestyle with constant multitasking and a lack of downtime.\nC. The activities reflect a chaotic and unplanned lifestyle with no clear routine or habits.\nD. The activities depict a balanced lifestyle that values both relaxation and cleanliness.\nE. The actions portray a highly structured and regimented lifestyle focused on productivity and order.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "D"}, {"from": "human", "value": "What can be inferred about the individual's state of mind and priorities based on their actions and the cleanliness of their environment?\nA. The individual is deeply engrossed in creative projects, with cleanliness being a secondary concern.\nB. The individual is primarily focused on achieving peak physical fitness, with little regard for their immediate surroundings.\nC. The individual values a peaceful and organized environment.\nD. The person is preoccupied with planning and organization, often neglecting personal relaxation.\nE. The individual is mainly driven by social interactions, placing minimal importance on personal space and cleanliness.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "C"}], "video": ["media/000113_P_13_1290.00_1350.00.mp4"]}
+{"id": "P_05_long_60_1", "conversations": [{"from": "human", "value": "<video>\nAre there books on the shelf in the room?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Is the room brightly lit?"}, {"from": "gpt", "value": "No."}], "video": ["media/000156_P_05_1410.00_1440.00.mp4"]}
+{"id": "P_15_short_23_1", "conversations": [{"from": "human", "value": "<video>\nAm I using a tablet?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "Is the room organized or cluttered?"}, {"from": "gpt", "value": "Cluttered."}], "video": ["media/000284_P_15_510.00_540.00.mp4"]}
+{"id": "P_03_long_39_1", "conversations": [{"from": "human", "value": "<video>\nAm I using a brush to wash dishes?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "Are the dishes dirty after washing?"}, {"from": "gpt", "value": "No."}], "video": ["media/000444_P_03_937.00_960.00.mp4"]}
+{"id": "P_13_short_80_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I gathering from around the house?"}, {"from": "gpt", "value": "Laundry."}, {"from": "human", "value": "What is my intent in gathering the laundry?"}, {"from": "gpt", "value": "To collect all the dirty laundry to prepare it for washing."}, {"from": "human", "value": "What might be my next step after gathering the laundry?"}, {"from": "gpt", "value": "To take the collected laundry to the washing machine."}, {"from": "human", "value": "What other activities indicate that I have been tidying up the kitchen area as well?"}, {"from": "gpt", "value": "There are dishes in the sink and cleaning supplies on the counter."}, {"from": "human", "value": "Is the kitchen counter completely empty in this segment?"}, {"from": "gpt", "value": "No."}], "video": ["media/000472_P_13_1650.00_1680.00.mp4"]}
+{"id": "P_02_short_136_1", "conversations": [{"from": "human", "value": "<video>\nIs the room dark?"}, {"from": "gpt", "value": "This information is not provided."}, {"from": "human", "value": "Am I outside?"}, {"from": "gpt", "value": "No."}], "video": ["media/000731_P_02_2430.00_2444.00.mp4"]}
+{"id": "P_02_long_58_1", "conversations": [{"from": "human", "value": "<video>\nAm I moving dishes to the living room?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "What might be my next step?"}, {"from": "gpt", "value": "To put the dishes in their proper places."}], "video": ["media/001069_P_02_1350.00_1377.00.mp4"]}
+{"id": "P_13_short_81_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I gathering from around the house?"}, {"from": "gpt", "value": "Laundry."}, {"from": "human", "value": "What is my intent while gathering laundry?"}, {"from": "gpt", "value": "To collect all the dirty clothes and linens for washing."}, {"from": "human", "value": "Am I collecting clean clothes?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "Where am I gathering laundry from?"}, {"from": "gpt", "value": "Around the house."}, {"from": "human", "value": "What will I do with the laundry after gathering it?"}, {"from": "gpt", "value": "Wash it."}], "video": ["media/001087_P_13_1680.00_1688.00.mp4"]}
+{"id": "P_03_short_66_1", "conversations": [{"from": "human", "value": "<video>\nIs the room brightly lit?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Does the bright lighting suggest it is daytime?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/001141_P_03_1321.00_1350.00.mp4"]}
+{"id": "P_13_long_60_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing in the video?"}, {"from": "gpt", "value": "Doing laundry."}, {"from": "human", "value": "What is my intent in this video?"}, {"from": "gpt", "value": "To ensure all the dirty clothes are cleaned properly."}, {"from": "human", "value": "Where am I in the video?"}, {"from": "gpt", "value": "In a room with a window that lets in natural light."}, {"from": "human", "value": "What is the first step I take?"}, {"from": "gpt", "value": "Gathering the laundry items."}, {"from": "human", "value": "Is there a cleaning tool nearby?"}, {"from": "gpt", "value": "Yes, a mop or similar cleaning tool."}], "video": ["media/001143_P_13_1590.00_1620.00.mp4"]}
+{"id": "P_03_long_42_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I making in this scene?"}, {"from": "gpt", "value": "Tea."}, {"from": "human", "value": "What might I do after making the tea?"}, {"from": "gpt", "value": "Pour the tea into a cup and add desired ingredients like sugar or milk."}, {"from": "human", "value": "What is heating on the stove?"}, {"from": "gpt", "value": "The teapot."}, {"from": "human", "value": "Are the cups prepared on the counter?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "What kind of atmosphere does the kitchen have?"}, {"from": "gpt", "value": "A cozy and warm atmosphere."}], "video": ["media/001344_P_03_1020.00_1050.00.mp4"]}
+{"id": "P_03_long_6_0", "conversations": [{"from": "human", "value": "<video>\nHow do the activities in the video reflect a structured approach to personal and household cleanliness?\nA. The activities illustrate a methodical sequence and meticulous attention to detail in cleaning.\nB. The actions imply the individual is more interested in spending time in the bathroom than achieving cleanliness.\nC. The activities seem more about gathering cleaning supplies rather than utilizing them for cleaning purposes.\nD. The activities appear to be spontaneous and lack a clear order, reflecting a chaotic approach to cleanliness.\nE. The individual is mainly focused on organizing personal hygiene products without engaging in actual cleaning.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "A"}, {"from": "human", "value": "What overarching theme can be derived from the individual tasks performed in the video?\nA. The overarching theme is the individual's fascination with personal care products.\nB. The primary theme centers around preparing for an important event or occasion.\nC. The theme is the individual's struggle to locate the appropriate cleaning supplies.\nD. The main theme is upholding a high standard of cleanliness and organization.\nE. The tasks suggest a constant state of anxiety about achieving cleanliness.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "D"}, {"from": "human", "value": "Considering the actions depicted, what can be inferred about the person's priorities regarding cleanliness and organization?\nA. The individual is more focused on the aesthetics of the bathroom than on actual cleanliness.\nB. The person places a high priority on thorough cleaning and maintaining an organized space.\nC. The priority seems to be on acquiring new cleaning supplies over using them effectively.\nD. The person seems more concerned with quickly drying items rather than cleaning them thoroughly.\nE. The individual's main priority appears to be spending as much time as possible in the bathroom.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "B"}, {"from": "human", "value": "How do the different environments in the video contribute to the overall narrative of the tasks being performed?\nA. The well-organized environments highlight the person's dedication to cleanliness.\nB. The settings indicate a disconnect between the person's actions and their cleanliness goals.\nC. The various environments suggest the person is frequently distracted and unable to focus on one task.\nD. The different environments create a chaotic narrative, showing inconsistency in cleaning habits.\nE. The environments portray a luxurious lifestyle that overshadows the cleaning tasks.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "A"}, {"from": "human", "value": "What key steps are involved in the person's routine, and how do they collectively ensure thorough cleanliness?\nA. The key steps focus more on organizing products than on actual cleaning.\nB. The individual engages in a repetitive cycle of cleaning without achieving thorough cleanliness.\nC. The routine involves random acts of cleaning mixed with periods of rest.\nD. The routine includes a step-by-step process of drying, washing, and organizing.\nE. The routine emphasizes quick surface cleaning rather than thorough cleaning steps.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "D"}], "video": ["media/001356_P_03_255.00_300.00.mp4"]}
+{"id": "P_02_short_6_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing in this scene?"}, {"from": "gpt", "value": "Drying my hands and face."}, {"from": "human", "value": "What is my intent while drying my hands and face?"}, {"from": "gpt", "value": "To remove any remaining moisture."}, {"from": "human", "value": "What do I plan to do next?"}, {"from": "gpt", "value": "Put away my toiletries and leave the bathroom."}, {"from": "human", "value": "Is the towel easily accessible?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Does the bathroom appear organized?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/001438_P_02_94.00_98.00.mp4"]}
+{"id": "P_15_short_3_1", "conversations": [{"from": "human", "value": "<video>\nIs the room disorganized at the start of this task?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Am I folding clothes in this specific part of the video?"}, {"from": "gpt", "value": "No."}], "video": ["media/001637_P_15_60.00_70.00.mp4"]}
+{"id": "P_02_long_12_0", "conversations": [{"from": "human", "value": "<video>\nHow do the surroundings in the two segments reflect the individual's approach to multitasking and the urgency of their actions?\nA. The settings illustrate that the person is extremely thorough and prepares every detail meticulously, no matter the location.\nB. The bathroom environment in the first part implies urgency and quick thinking, whereas the kitchen environment in the second part suggests a more planned and systematic approach.\nC. The environments suggest that the person is mainly concentrating on cleaning tasks rather than on food preparation.\nD. The individual's multitasking method remains consistent in both segments, showing no variation in urgency or surroundings.\nE. The surroundings indicate that the individual lacks a sense of urgency and is leisurely going about their food preparation.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "B"}, {"from": "human", "value": "Compare the individual's approach to food preparation in both segments and summarize the key differences and similarities.\nA. The first segment focuses on hydration, while the second segment is solely about preparing food.\nB. The food preparation in the first segment is completely different from that in the second, with no common approach.\nC. In both segments, the individual is preparing food, but the first segment appears more spontaneous, whereas the second is more organized and planned.\nD. The person follows intricate recipes and detailed steps in both segments, indicating a methodical approach throughout.\nE. The main difference is that the individual makes a hot dish in the first segment and a cold snack in the second.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "C"}, {"from": "human", "value": "What can be inferred about the individual's priorities based on their simultaneous actions and location in the video?\nA. The individual places a high priority on cleanliness and hygiene, as shown by the presence of personal care items.\nB. The main priority for the person is staying hydrated, as indicated by the continuous consumption of water.\nC. The person focuses mainly on using a particular set of ingredients for their snack, showing ingredient preference.\nD. The individual's top priority is to reflect on their actions while looking in the mirror and preparing food.\nE. The person prioritizes effectiveness and practicality, even if it involves using unconventional locations to get things done.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "E"}, {"from": "human", "value": "Identify the most crucial aspect of the individual's behavior that indicates their primary objective in the video.\nA. The most crucial aspect is the individual's commitment to maintaining a clean and orderly preparation area.\nB. The vital behavior is the individual’s attention to the range of ingredients used in the snack preparation.\nC. The key aspect is the individual's aim to efficiently prepare a snack while also ensuring they stay hydrated.\nD. The main objective is to display their multitasking abilities by using both the bathroom and the kitchen.\nE. The principal goal is to identify the most convenient place for food preparation, irrespective of the setting.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "C"}, {"from": "human", "value": "How does the change in location between the two segments affect the overall context of the individual's actions and intent?\nA. The shift from the bathroom to the kitchen suggests an increased emphasis on personal hygiene during food preparation.\nB. The change in location does not significantly affect the individual's actions or intentions, which remain constant throughout.\nC. The move from the bathroom to the kitchen highlights the individual's growing unease with multitasking in less conventional places.\nD. The change in surroundings indicates that the individual appreciates variety in their food preparation settings.\nE. The transition from the bathroom to the kitchen implies a move from a makeshift approach to a more conventional and organized method.\nAnswer with the option's letter from the given choices directly."}, {"from": "gpt", "value": "E"}], "video": ["media/002086_P_02_570.00_630.00.mp4"]}
+{"id": "P_03_long_13_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing with the towel?"}, {"from": "gpt", "value": "Drying my hands/face."}, {"from": "human", "value": "What is my intent while using the towel?"}, {"from": "gpt", "value": "To complete my cleaning routine."}, {"from": "human", "value": "How is the bathroom organized?"}, {"from": "gpt", "value": "With items like mouthwash, soap, and other personal care items."}, {"from": "human", "value": "What might be my next step after drying my hands/face?"}, {"from": "gpt", "value": "Leaving the bathroom or proceeding with another personal care task."}, {"from": "human", "value": "Is the bathroom messy or well-organized?"}, {"from": "gpt", "value": "Well-organized."}], "video": ["media/002196_P_03_255.00_260.00.mp4"]}
+{"id": "P_02_short_124_1", "conversations": [{"from": "human", "value": "<video>\nIs the workspace disorganized?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "What is the next step after spreading peanut butter?"}, {"from": "gpt", "value": "Eating the prepared dish."}], "video": ["media/002378_P_02_2210.00_2230.00.mp4"]}
+{"id": "P_03_long_53_1", "conversations": [{"from": "human", "value": "<video>\nIs it likely to be daytime or nighttime?"}, {"from": "gpt", "value": "Daytime."}, {"from": "human", "value": "Are there various items on the kitchen counter?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/002443_P_03_1321.00_1350.00.mp4"]}
+{"id": "P_12_short_56_1", "conversations": [{"from": "human", "value": "<video>\nDoes the scene suggest I am working or studying?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Is the desk empty?"}, {"from": "gpt", "value": "No."}], "video": ["media/002627_P_12_840.00_844.00.mp4"]}
+{"id": "P_12_long_2_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing in the video?"}, {"from": "gpt", "value": "Brushing my teeth."}, {"from": "human", "value": "What is the purpose of my activity?"}, {"from": "gpt", "value": "To maintain oral hygiene."}, {"from": "human", "value": "What might I do next after brushing my teeth?"}, {"from": "gpt", "value": "Wash my hands or face."}, {"from": "human", "value": "What items are visible on the counter in front of me?"}, {"from": "gpt", "value": "Toothbrush and toothpaste."}, {"from": "human", "value": "Is the water running from the tap while I brush my teeth?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/002783_P_12_30.00_58.00.mp4"]}

egoit_Ego4D/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "egoit_Ego4D",
+  "modality": [
+    "video"
+  ],
+  "category": "embodied_video_vqa",
+  "total": 472,
+  "sampled": 20
+}

egoit_EgoProceL/sampled.jsonl

@@ -0,0 +1,20 @@
+{"id": "0002_short_463_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I plugging into the partial model?"}, {"from": "gpt", "value": "A rod."}, {"from": "human", "value": "What is my intent for plugging the rod into the partial model?"}, {"from": "gpt", "value": "To secure the rod in place as part of the assembly process."}, {"from": "human", "value": "What might I do next after plugging the rod into the partial model?"}, {"from": "gpt", "value": "Ensure that the rod is properly aligned and attached."}, {"from": "human", "value": "What color is the partial model I am working on?"}, {"from": "gpt", "value": "Red."}, {"from": "human", "value": "Am I carefully inserting the rod into the model?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/000003_0002_1305.87_1306.47.mp4"]}
+{"id": "0012_short_242_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing?"}, {"from": "gpt", "value": "Taking a screw."}, {"from": "human", "value": "What is my intent for taking the screw?"}, {"from": "gpt", "value": "To use it for the next assembly step."}, {"from": "human", "value": "What do I plan to do next?"}, {"from": "gpt", "value": "Check the booklet again to make sure I am using the correct screw."}, {"from": "human", "value": "What color is the screw I am picking up?"}, {"from": "gpt", "value": "Red."}, {"from": "human", "value": "Am I preparing to use the screw in the assembly?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/001425_0012_386.13_386.93.mp4"]}
+{"id": "0012_short_314_1", "conversations": [{"from": "human", "value": "<video>\nAm I ensuring the rim is oriented correctly?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Is the rim piece square-shaped?"}, {"from": "gpt", "value": "No."}], "video": ["media/001721_0012_544.60_545.13.mp4"]}
+{"id": "S07_Eggs_7150991-1466_long_7_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I pouring into the pan?"}, {"from": "gpt", "value": "The mixture."}, {"from": "human", "value": "What is the intent of pouring the mixture into the pan?"}, {"from": "gpt", "value": "To cook the mixture, possibly making an omelet or scrambled eggs."}, {"from": "human", "value": "What tool am I using to guide and spread the mixture?"}, {"from": "gpt", "value": "A fork."}, {"from": "human", "value": "What might be my next step after pouring the mixture into the pan?"}, {"from": "gpt", "value": "To monitor the cooking process, occasionally stirring or flipping the mixture."}, {"from": "human", "value": "What color is the bowl from which I am pouring the mixture?"}, {"from": "gpt", "value": "White."}], "video": ["media/001750_S07_Eggs_7150991-1466_133.98_143.14.mp4"]}
+{"id": "0002_long_289_0", "conversations": [{"from": "human", "value": "<video>\nWhat do I do with the rod?"}, {"from": "gpt", "value": "Plug it into the model and pull it to ensure it is secure."}, {"from": "human", "value": "What is my intent with the rod?"}, {"from": "gpt", "value": "To finalize the connection of this part of the model."}, {"from": "human", "value": "What might my next step be after securing the rod?"}, {"from": "gpt", "value": "To add any final components or tighten the existing connections."}, {"from": "human", "value": "Am I carefully ensuring each connection is secure?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Is the model starting to look more complete?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/002350_0002_1346.20_1358.40.mp4"]}
+{"id": "0012_short_238_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing with the objects?"}, {"from": "gpt", "value": "Aligning them."}, {"from": "human", "value": "What is my intent with aligning the objects?"}, {"from": "gpt", "value": "To ensure the components fit together correctly and are ready for screwing."}, {"from": "human", "value": "What might be my next step after aligning the objects?"}, {"from": "gpt", "value": "To screw the screw with the screwdriver to secure the parts."}, {"from": "human", "value": "How am I positioning the parts of the model?"}, {"from": "gpt", "value": "Carefully, to ensure proper alignment."}, {"from": "human", "value": "Is the instruction booklet open nearby?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/002465_0012_381.53_382.27.mp4"]}
+{"id": "27.tent.071119.gopro_long_10_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I inserting into the ground?"}, {"from": "gpt", "value": "Stakes."}, {"from": "human", "value": "What might be my next step after inserting the stakes?"}, {"from": "gpt", "value": "Ensuring all sides of the tent are evenly staked down."}, {"from": "human", "value": "What is the purpose of securing the tent with stakes?"}, {"from": "gpt", "value": "To create a stable and secure shelter."}, {"from": "human", "value": "What is my surroundings like?"}, {"from": "gpt", "value": "A grassy area covered with fallen leaves."}, {"from": "human", "value": "What am I holding in one hand?"}, {"from": "gpt", "value": "A stake."}], "video": ["media/002489_27.tent.071119.gopro_519.24_542.71.mp4"]}
+{"id": "0002_short_499_1", "conversations": [{"from": "human", "value": "<video>\nAm I holding the screwdriver in this timestamp?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "Is the assembly complete?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/002639_0002_1402.93_1403.60.mp4"]}
+{"id": "0002_short_3_1", "conversations": [{"from": "human", "value": "<video>\nIs the booklet being used for cooking instructions?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "Am I meticulously looking at the illustrations in the booklet?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/003326_0002_7.80_8.47.mp4"]}
+{"id": "04.tent.061417.gopro_short_17_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing in this video?"}, {"from": "gpt", "value": "Inserting a stake into the ground to secure a tent."}, {"from": "human", "value": "What is my intent when inserting the stake?"}, {"from": "gpt", "value": "To ensure that the tent remains stable and anchored, especially in case of wind or other disturbances."}, {"from": "human", "value": "How am I pushing the stake into the ground?"}, {"from": "gpt", "value": "Using both hands to push the stake firmly into the grass."}, {"from": "human", "value": "What might my next step be?"}, {"from": "gpt", "value": "To check the tension of the ropes and adjust them if necessary."}, {"from": "human", "value": "What am I wearing?"}, {"from": "gpt", "value": "A leopard-print garment."}], "video": ["media/003521_04.tent.061417.gopro_431.01_438.05.mp4"]}
+{"id": "0013_short_255_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I plugging into the newly aligned parts?"}, {"from": "gpt", "value": "Another rod."}, {"from": "human", "value": "What is my intent while plugging the rod?"}, {"from": "gpt", "value": "To continue securing the model piece by piece."}, {"from": "human", "value": "What might be my next step after plugging the rod?"}, {"from": "gpt", "value": "To check the stability of the assembled parts and make any necessary adjustments."}, {"from": "human", "value": "Am I inserting the rod carefully?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Is my approach methodical in this step?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/003619_0013_522.27_523.40.mp4"]}
+{"id": "0002_short_383_1", "conversations": [{"from": "human", "value": "<video>\nIs the screwdriver intended for loosening screws?"}, {"from": "gpt", "value": "No."}, {"from": "human", "value": "Am I using the screwdriver to hammer nails?"}, {"from": "gpt", "value": "No."}], "video": ["media/003622_0002_1009.80_1010.67.mp4"]}
+{"id": "0002_long_216_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I checking to confirm the next steps in the assembly process?"}, {"from": "gpt", "value": "The booklet."}, {"from": "human", "value": "What is my intent in this segment?"}, {"from": "gpt", "value": "To ensure I am following the instructions correctly to avoid mistakes."}, {"from": "human", "value": "What might be my next step?"}, {"from": "gpt", "value": "To take the partial model and plug another screw into it, as indicated by the instructions."}, {"from": "human", "value": "Am I comparing the parts in my hands with the images in the booklet?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "What am I carefully comparing to ensure accuracy?"}, {"from": "gpt", "value": "The parts in my hands with the images in the booklet."}], "video": ["media/003809_0002_901.53_926.60.mp4"]}
+{"id": "0012_short_278_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I plugging the screw into?"}, {"from": "gpt", "value": "One of the red plastic parts I am holding."}, {"from": "human", "value": "What is my intent in this step?"}, {"from": "gpt", "value": "To secure the parts together as per the instruction manual."}, {"from": "human", "value": "What might my next step be?"}, {"from": "gpt", "value": "To ensure the screw is tightly fitted, possibly by using a washer or bolt."}, {"from": "human", "value": "Are the red plastic parts being aligned?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Is the screw being fitted into a pre-drilled hole?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/004488_0012_450.67_450.93.mp4"]}
+{"id": "0012_short_225_1", "conversations": [{"from": "human", "value": "<video>\nIs the booklet showing irrelevant information?"}, {"from": "gpt", "value": "No"}, {"from": "human", "value": "What is the booklet open to?"}, {"from": "gpt", "value": "The relevant page with assembly instructions."}], "video": ["media/004662_0012_358.13_360.60.mp4"]}
+{"id": "0012_short_215_0", "conversations": [{"from": "human", "value": "<video>\nWhat do I put into the partial model?"}, {"from": "gpt", "value": "The screw."}, {"from": "human", "value": "What is my intent after putting the screw into the partial model?"}, {"from": "gpt", "value": "To secure the components together to ensure stability in the structure."}, {"from": "human", "value": "What might be my next step after inserting the screw?"}, {"from": "gpt", "value": "To check the booklet to confirm if the assembly is correct."}, {"from": "human", "value": "Are there various red and grey plastic parts around me?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Is the booklet closed after I put the screw into the model?"}, {"from": "gpt", "value": "No."}], "video": ["media/004751_0012_343.87_344.80.mp4"]}
+{"id": "05.tent.061417.gopro_long_7_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I doing with the support tab in this scene?"}, {"from": "gpt", "value": "Inserting it into the tent pole."}, {"from": "human", "value": "What is my intent in this scene?"}, {"from": "gpt", "value": "To stabilize the tent by properly assembling its framework."}, {"from": "human", "value": "What might be my next step after securing the support tab into the tent pole?"}, {"from": "gpt", "value": "To secure the other ends of the tent poles into their respective slots."}, {"from": "human", "value": "What am I wearing in this scene?"}, {"from": "gpt", "value": "A burgundy long-sleeve shirt, jeans, and brown shoes."}, {"from": "human", "value": "What is the color of the tent?"}, {"from": "gpt", "value": "Blue and black."}], "video": ["media/004969_05.tent.061417.gopro_201.47_254.32.mp4"]}
+{"id": "27.tent.071119.gopro_short_1_0", "conversations": [{"from": "human", "value": "<video>\nWhat color is the tent bag?"}, {"from": "gpt", "value": "Blue and gray."}, {"from": "human", "value": "What am I wearing?"}, {"from": "gpt", "value": "A black jacket and ripped jeans."}, {"from": "human", "value": "Where am I most likely setting up the tent?"}, {"from": "gpt", "value": "In a grassy area with colorful autumn leaves."}, {"from": "human", "value": "What is my next likely step after unzipping the bag?"}, {"from": "gpt", "value": "Fully remove the tent from the bag and begin assembling it."}, {"from": "human", "value": "Am I indoors or outdoors?"}, {"from": "gpt", "value": "Outdoors."}], "video": ["media/005042_27.tent.071119.gopro_49.73_54.59.mp4"]}
+{"id": "0002_short_117_1", "conversations": [{"from": "human", "value": "<video>\nIs the structure looking more complete with each bolt?"}, {"from": "gpt", "value": "Yes."}, {"from": "human", "value": "Will I leave some bolts loose?"}, {"from": "gpt", "value": "No."}], "video": ["media/005363_0002_255.67_258.60.mp4"]}
+{"id": "0013_short_296_0", "conversations": [{"from": "human", "value": "<video>\nWhat am I aligning in front of me?"}, {"from": "gpt", "value": "The objects (parts) for assembly."}, {"from": "human", "value": "What is my intent when aligning the objects?"}, {"from": "gpt", "value": "To create a clear workspace to make it easier to find the parts I need for the construction."}, {"from": "human", "value": "What might be my next step after aligning the objects?"}, {"from": "gpt", "value": "To start connecting these parts according to the instructions to build the structure."}, {"from": "human", "value": "What components can I see laid out in front of me?"}, {"from": "gpt", "value": "Several red and grey components."}, {"from": "human", "value": "Is the instruction booklet open on the table?"}, {"from": "gpt", "value": "Yes."}], "video": ["media/005534_0013_619.40_619.93.mp4"]}

egoplan_val_converted/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "egoplan_val_converted",
+  "modality": [
+    "image"
+  ],
+  "category": "video_vqa",
+  "total": 3353,
+  "sampled": 20
+}

egoplan_val_converted/sampled.jsonl

@@ -0,0 +1,20 @@
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to pick up knife, cut courgette, search for peeler in drawer and pick it up?\n\nA. pick up knife\n\nB. open drawer\n\nC. pick up peeler\n\nD. search for peeler\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "C"}], "image": ["media/000277_kf0.jpg", "media/000277_kf1.jpg", "media/000277_kf2.jpg", "media/000277_kf3.jpg", "media/000277_kf4.jpg", "media/000277_kf5.jpg", "media/000277_kf6.jpg", "media/000277_kf7.jpg"], "id": "21"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to wash the knife and cutting board?\n\nA. turn on tap\n\nB. wash cutting board\n\nC. rinse knife\n\nD. close tap\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "B"}], "image": ["media/000553_kf0.jpg", "media/000553_kf1.jpg", "media/000553_kf2.jpg", "media/000553_kf3.jpg", "media/000553_kf4.jpg", "media/000553_kf5.jpg", "media/000553_kf6.jpg", "media/000553_kf7.jpg"], "id": "617"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to pour oil into pan and close oil bottle?\n\nA. pour oil into pan\n\nB. take oil bottle\n\nC. close oil bottle\n\nD. open oil bottle\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "A"}], "image": ["media/000565_kf0.jpg", "media/000565_kf1.jpg", "media/000565_kf2.jpg", "media/000565_kf3.jpg", "media/000565_kf4.jpg", "media/000565_kf5.jpg", "media/000565_kf6.jpg", "media/000565_kf7.jpg"], "id": "709"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to clean and store pot?\n\nA. put pot in drawer\n\nB. close drawer\n\nC. drain pot\n\nD. clean pot\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "D"}], "image": ["media/000594_kf0.jpg", "media/000594_kf1.jpg", "media/000594_kf2.jpg", "media/000594_kf3.jpg", "media/000594_kf4.jpg", "media/000594_kf5.jpg", "media/000594_kf6.jpg", "media/000594_kf7.jpg"], "id": "770"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to clean up kitchen?\n\nA. pick up cloth\n\nB. open food bin\n\nC. rinse hands\n\nD. shake hands\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "A"}], "image": ["media/000647_kf0.jpg", "media/000647_kf1.jpg", "media/000647_kf2.jpg", "media/000647_kf3.jpg", "media/000647_kf4.jpg", "media/000647_kf5.jpg", "media/000647_kf6.jpg", "media/000647_kf7.jpg"], "id": "599"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to take saucepan and lid, put saucepan and close cupboard?\n\nA. take saucepan lid\n\nB. close cupboard\n\nC. put saucepan\n\nD. take saucepan\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "B"}], "image": ["media/000828_kf0.jpg", "media/000828_kf1.jpg", "media/000828_kf2.jpg", "media/000828_kf3.jpg", "media/000828_kf4.jpg", "media/000828_kf5.jpg", "media/000828_kf6.jpg", "media/000828_kf7.jpg"], "id": "860"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to put sandwiches and crisps in tupperware?\n\nA. close cupboard\n\nB. open cupboard\n\nC. pick up lid\n\nD. pick up sandwiches\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "D"}], "image": ["media/001265_kf0.jpg", "media/001265_kf1.jpg", "media/001265_kf2.jpg", "media/001265_kf3.jpg", "media/001265_kf4.jpg", "media/001265_kf5.jpg", "media/001265_kf6.jpg", "media/001265_kf7.jpg"], "id": "1250"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to scoop up food from sink and wash utensils?\n\nA. take paper from paper roll\n\nB. wash spoon\n\nC. scoop up food from sink\n\nD. pick up paper roll\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "C"}], "image": ["media/001300_kf0.jpg", "media/001300_kf1.jpg", "media/001300_kf2.jpg", "media/001300_kf3.jpg", "media/001300_kf4.jpg", "media/001300_kf5.jpg", "media/001300_kf6.jpg", "media/001300_kf7.jpg"], "id": "1189"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to clean up the kitchen and put away ingredients?\n\nA. close bottle\n\nB. fold mozzarella bag\n\nC. put down beer\n\nD. close fridge\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "B"}], "image": ["media/001761_kf0.jpg", "media/001761_kf1.jpg", "media/001761_kf2.jpg", "media/001761_kf3.jpg", "media/001761_kf4.jpg", "media/001761_kf5.jpg", "media/001761_kf6.jpg", "media/001761_kf7.jpg"], "id": "1874"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to prepare the spicy bean sauce?\n\nA. open spicy bean sauce\n\nB. put down spicy bean sauce\n\nC. put down yoghurt\n\nD. pick up spicy bean sauce\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "C"}], "image": ["media/002145_kf0.jpg", "media/002145_kf1.jpg", "media/002145_kf2.jpg", "media/002145_kf3.jpg", "media/002145_kf4.jpg", "media/002145_kf5.jpg", "media/002145_kf6.jpg", "media/002145_kf7.jpg"], "id": "2082"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to prepare the dish?\n\nA. take bowl\n\nB. move dish\n\nC. put dishes\n\nD. take fork\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "B"}], "image": ["media/002183_kf0.jpg", "media/002183_kf1.jpg", "media/002183_kf2.jpg", "media/002183_kf3.jpg", "media/002183_kf4.jpg", "media/002183_kf5.jpg", "media/002183_kf6.jpg", "media/002183_kf7.jpg"], "id": "2264"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to clean the sink and counter top?\n\nA. pick up degreaser\n\nB. move coffee pod\n\nC. move tap\n\nD. pick up cloth\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "B"}], "image": ["media/002217_kf0.jpg", "media/002217_kf1.jpg", "media/002217_kf2.jpg", "media/002217_kf3.jpg", "media/002217_kf4.jpg", "media/002217_kf5.jpg", "media/002217_kf6.jpg", "media/002217_kf7.jpg"], "id": "2330"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to prepare the bread dough and preheat the pan?\n\nA. take cling film off bread dough\n\nB. open oven\n\nC. open cupboard\n\nD. put cast iron pan in oven\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "B"}], "image": ["media/002282_kf0.jpg", "media/002282_kf1.jpg", "media/002282_kf2.jpg", "media/002282_kf3.jpg", "media/002282_kf4.jpg", "media/002282_kf5.jpg", "media/002282_kf6.jpg", "media/002282_kf7.jpg"], "id": "2251"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to dispose of trash?\n\nA. pick up bag\n\nB. put down cloth\n\nC. fold bag\n\nD. close garbage can\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "D"}], "image": ["media/002313_kf0.jpg", "media/002313_kf1.jpg", "media/002313_kf2.jpg", "media/002313_kf3.jpg", "media/002313_kf4.jpg", "media/002313_kf5.jpg", "media/002313_kf6.jpg", "media/002313_kf7.jpg"], "id": "2218"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to wash the chopsticks?\n\nA. touch strainer\n\nB. open tap\n\nC. wash chopsticks\n\nD. put down bottle soap\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "D"}], "image": ["media/002687_kf0.jpg", "media/002687_kf1.jpg", "media/002687_kf2.jpg", "media/002687_kf3.jpg", "media/002687_kf4.jpg", "media/002687_kf5.jpg", "media/002687_kf6.jpg", "media/002687_kf7.jpg"], "id": "2736"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to prepare utensils?\n\nA. pull drawer\n\nB. place chopsticks on tray\n\nC. open drawer\n\nD. pick up chopsticks\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "C"}], "image": ["media/002801_kf0.jpg", "media/002801_kf1.jpg", "media/002801_kf2.jpg", "media/002801_kf3.jpg", "media/002801_kf4.jpg", "media/002801_kf5.jpg", "media/002801_kf6.jpg", "media/002801_kf7.jpg"], "id": "2898"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next for getting a plate?\n\nA. close cabinet\n\nB. open cabinet\n\nC. pick up plate\n\nD. put down plate on table\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "D"}], "image": ["media/002820_kf0.jpg", "media/002820_kf1.jpg", "media/002820_kf2.jpg", "media/002820_kf3.jpg", "media/002820_kf4.jpg", "media/002820_kf5.jpg", "media/002820_kf6.jpg", "media/002820_kf7.jpg"], "id": "2613"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next for cleaning glasses and cups?\n\nA. put cup on shelf\n\nB. pick up glass\n\nC. pick up cup\n\nD. put glass on shelf\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "D"}], "image": ["media/002888_kf0.jpg", "media/002888_kf1.jpg", "media/002888_kf2.jpg", "media/002888_kf3.jpg", "media/002888_kf4.jpg", "media/002888_kf5.jpg", "media/002888_kf6.jpg", "media/002888_kf7.jpg"], "id": "2825"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to store leftover food and cabbage in the cabinet and fridge?\n\nA. pick up bottle of salad\n\nB. open fridge\n\nC. insert cabbage into fridge\n\nD. pick up plastic tin\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "C"}], "image": ["media/003154_kf0.jpg", "media/003154_kf1.jpg", "media/003154_kf2.jpg", "media/003154_kf3.jpg", "media/003154_kf4.jpg", "media/003154_kf5.jpg", "media/003154_kf6.jpg", "media/003154_kf7.jpg"], "id": "3304"}
+{"conversations": [{"from": "human", "value": "<image><image><image><image><image><image><image><image>\nSelect the best answer to the following multiple-choice question based on the video.\n\nConsidering the progress shown in the video and my current observation in the last frame, what action should I take next in order to clean the knife and chopping board?\n\nA. pick up chopping board\n\nB. put down chopping board\n\nC. pick up knife\n\nD. wash hands\n\nRespond with only the letter (A, B, C, or D) of the correct option."}, {"from": "gpt", "value": "C"}], "image": ["media/003277_kf0.jpg", "media/003277_kf1.jpg", "media/003277_kf2.jpg", "media/003277_kf3.jpg", "media/003277_kf4.jpg", "media/003277_kf5.jpg", "media/003277_kf6.jpg", "media/003277_kf7.jpg"], "id": "3235"}

eo_data_episode_caption/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "eo_data_episode_caption",
+  "modality": [
+    "image"
+  ],
+  "category": "embodied_episode_caption",
+  "total": 84832,
+  "sampled": 20
+}

eo_data_object_referring_qa/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "eo_data_object_referring_qa",
+  "modality": [
+    "image"
+  ],
+  "category": "embodied_object_referring",
+  "total": 176735,
+  "sampled": 20
+}

eo_data_object_referring_qa/sampled.jsonl

@@ -0,0 +1,20 @@
+{"id": "qa-object_referring_qa_26300", "conversations": [{"from": "human", "value": "<image-1>Please locate all items in this image and list their bounding boxes. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<539><369><588><494>, <179><565><305><720>, <104><696><182><798>, <588><470><659><577>, <578><536><620><631>, <234><131><286><214>, <227><143><247><196>, <130><36><237><286>]</box>"}], "image": ["media/026287_00026300.jpg"]}
+{"id": "qa-object_referring_qa_31230", "conversations": [{"from": "human", "value": "<image-1>The task is to identify each object present in the image. For each, output. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<519><488><578><552>, <198><119><308><234>, <477><254><591><317>, <172><595><256><687>]</box>"}], "image": ["media/031217_00031230.jpg"]}
+{"id": "qa-object_referring_qa_36210", "conversations": [{"from": "human", "value": "<image-1> list all items from the image, ensuring each has its bounding boxes specified. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<669><201><857><638>, <410><299><471><496>, <445><210><490><317>, <471><152><521><263>, <588><482><648><580>, <605><549><674><670>, <617><629><695><786>, <667><643><755><777>]</box>"}], "image": ["media/036193_00036210.jpg"]}
+{"id": "qa-object_referring_qa_42435", "conversations": [{"from": "human", "value": "<image-1>The task is to identify each object present in the image. For each, output. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<104><266><379><587>, <566><282><821><496>, <407><194><580><496>]</box>"}], "image": ["media/042413_00042435.jpg"]}
+{"id": "qa-object_referring_qa_56243", "conversations": [{"from": "human", "value": "<image-1> list all items from the image, ensuring each has its bounding boxes specified. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<707><0><833><335>, <231><103><283><205>, <464><112><517><214>, <471><246><524><353>, <474><299><536><411>, <648><268><707><384>, <664><326><736><455>, <638><464><695><580>, <688><214><719><277>]</box>"}], "image": ["media/056218_00056243.jpg"]}
+{"id": "qa-object_referring_qa_60259", "conversations": [{"from": "human", "value": "<image-1>Detect all individual objects within the given image and output their corresponding. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<903><845><1000><901>, <175><234><308><373>, <429><250><503><365>, <620><274><695><373>, <649><63><737><274>]</box>"}], "image": ["media/060233_00060259.jpg"]}
+{"id": "qa-object_referring_qa_66554", "conversations": [{"from": "human", "value": "<image-1>Could you identify all distinct entities in the image and return their bounding boxes?. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<523><0><698><351>, <234><792><308><988>, <344><446><523><857>, <0><214><338><685>, <250><48><432><250>, <659><155><737><208>, <679><244><896><321>, <698><613><782><726>, <695><744><789><851>, <896><738><945><857>]</box>"}], "image": ["media/066526_00066554.jpg"]}
+{"id": "qa-object_referring_qa_80423", "conversations": [{"from": "human", "value": "<image-1>Could you identify all distinct entities in the image and return their bounding boxes?. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<494><471><604><657>, <464><536><747><786>, <220><143><917><911>, <74><289><982><1000>]</box>"}], "image": ["media/080387_00080423.jpg"]}
+{"id": "qa-object_referring_qa_88759", "conversations": [{"from": "human", "value": "<image-1>All objects in the image require identification with bounding boxes. Please ensure the output of all bounding boxes is. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<636><0><1000><456>, <91><353><367><651>, <510><274><737><488>, <455><468><558><651>]</box>"}], "image": ["media/088719_00088759.jpg"]}
+{"id": "qa-object_referring_qa_90310", "conversations": [{"from": "human", "value": "<image-1> what are all the objects in this image? Provide bounding boxes for them. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<681><762><1000><1000>, <223><810><374><1000>, <879><730><1000><980>, <390><337><679><988>, <832><87><1000><571>]</box>"}], "image": ["media/090270_00090310.jpg"]}
+{"id": "qa-object_referring_qa_104412", "conversations": [{"from": "human", "value": "<image-1>The task is to identify each object present in the image. For each, output. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<484><194><723><401>, <0><389><321><798>, <234><321><690><940>, <327><520><574><730>, <495><631><571><730>, <591><389><931><845>]</box>"}], "image": ["media/104367_00104412.jpg"]}
+{"id": "qa-object_referring_qa_105049", "conversations": [{"from": "human", "value": "<image-1>Please locate all items in this image and list their bounding boxes. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<673><651><964><1000>, <102><679><376><1000>, <648><647><750><762>]</box>"}], "image": ["media/105004_00105049.jpg"]}
+{"id": "qa-object_referring_qa_117120", "conversations": [{"from": "human", "value": "<image-1> list all items from the image, ensuring each has its bounding boxes specified. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<223><107><467><433>, <511><230><742><480>, <415><206><659><556>, <464><571><591><762>]</box>"}], "image": ["media/117072_00117120.jpg"]}
+{"id": "qa-object_referring_qa_126867", "conversations": [{"from": "human", "value": "<image-1>For every object depicted in the image, provide their bounding boxes. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<562><202><591><357>, <549><363><604><619>, <341><500><377><685>, <412><494><549><631>, <292><411><416><577>, <416><601><494><649>, <367><690><526><750>, <565><667><679><732>, <601><500><672><619>, <354><327><396><369>, <464><256><519><315>, <500><327><549><345>, <445><208><477><268>, <435><262><477><315>]</box>"}], "image": ["media/126814_00126867.jpg"]}
+{"id": "qa-object_referring_qa_133312", "conversations": [{"from": "human", "value": "<image-1>Detect all individual objects within the given image and output their corresponding. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<52><0><328><298>, <6><381><260><786>, <159><583><269><798>, <269><708><425><887>, <425><726><646><863>, <455><827><646><1000>, <675><738><828><1000>, <601><327><737><732>, <308><411><442><690>, <412><405><526><631>, <737><369><1000><792>]</box>"}], "image": ["media/133259_00133312.jpg"]}
+{"id": "qa-object_referring_qa_144391", "conversations": [{"from": "human", "value": "<image-1>Detect all individual objects within the given image and output their corresponding. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<140><639><462><865>, <538><567><775><794>, <247><516><563><948>, <118><472><923><1000>]</box>"}], "image": ["media/144330_00144391.jpg"]}
+{"id": "qa-object_referring_qa_145527", "conversations": [{"from": "human", "value": "<image-1>Detect all individual objects within the given image and output their corresponding. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<533><332><667><550>, <0><318><899><1000>, <205><425><756><921>, <295><796><405><843>, <333><746><509><804>, <518><711><610><804>, <479><843><634><886>, <247><575><286><711>, <0><0><241><196>]</box>"}], "image": ["media/145466_00145527.jpg"]}
+{"id": "qa-object_referring_qa_151490", "conversations": [{"from": "human", "value": "<image-1> what are all the objects in this image? Provide bounding boxes for them. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<321><161><521><554>, <298><771><363><868>, <554><557><640><668>, <357><196><530><461>, <48><204><238><379>, <39><321><202><921>, <182><168><869><943>]</box>"}], "image": ["media/151427_00151490.jpg"]}
+{"id": "qa-object_referring_qa_161032", "conversations": [{"from": "human", "value": "<image-1> what are all the objects in this image? Provide bounding boxes for them. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<0><147><254><580>, <750><152><964><554>, <397><371><679><710>]</box>"}], "image": ["media/160963_00161032.jpg"]}
+{"id": "qa-object_referring_qa_172901", "conversations": [{"from": "human", "value": "<image-1>Detect all individual objects within the given image and output their corresponding. Output bounding boxes directly."}, {"from": "gpt", "value": "<box>[<735><593><827><686>, <324><689><420><843>, <470><582><595><729>, <625><632><732><757>, <625><521><732><593>, <0><0><152><193>, <193><0><327><61>, <250><18><488><196>]</box>"}], "image": ["media/172827_00172901.jpg"]}

eo_data_subtask_qa/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "eo_data_subtask_qa",
+  "modality": [
+    "image"
+  ],
+  "category": "embodied_subtask_qa",
+  "total": 143159,
+  "sampled": 20
+}

eo_data_trajectory_qa/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "eo_data_trajectory_qa",
+  "modality": [
+    "image"
+  ],
+  "category": "embodied_trajectory_qa",
+  "total": 97191,
+  "sampled": 20
+}

eo_data_trajectory_qa/sampled.jsonl

@@ -0,0 +1,20 @@
+{"id": "qa-trajectory_qa_23957", "conversations": [{"from": "human", "value": "<image-1>To successfully pour a liquid from the blue and white can into the purple cup, describe the complete, multi-part trajectory the robotic arm's gripper must execute, starting from its current position and ending with the can upright again over the stovetop. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<821><782>, <770><694>, <663><643>, <516><659>, <452><750>, <504><810>]</trace>"}], "image": ["media/023957_00023957.jpg"]}
+{"id": "qa-trajectory_qa_24925", "conversations": [{"from": "human", "value": "<image-1>Outline the necessary robotic arm operations to achieve \"retrieve the object from the table\", ensuring you provide trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<635><599>, <587><643>, <548><683>, <508><718>, <472><750>, <437><782>]</trace>"}], "image": ["media/024925_00024925.jpg"]}
+{"id": "qa-trajectory_qa_26170", "conversations": [{"from": "human", "value": "<image-1>What is the optimal way for the robotic arm to move to achieve \"place the spatula on the purple towel\"? Deliver trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<230><353>, <294><393>, <369><440>, <433><484>, <496><528>, <524><563>]</trace>"}], "image": ["media/026170_00026170.jpg"]}
+{"id": "qa-trajectory_qa_26577", "conversations": [{"from": "human", "value": "<image-1>Provide the articulation sequence for the robotic arm to achieve \"move towards the other blue cube\", giving trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<278><325>, <313><393>, <345><460>, <381><528>, <417><599>, <853><341>]</trace>"}], "image": ["media/026577_00026577.jpg"]}
+{"id": "qa-trajectory_qa_27102", "conversations": [{"from": "human", "value": "<image-1>How can the robotic arm be guided to achieve \"place the silver pot on the lower left side of the table\"? List trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<321><341>, <266><504>, <238><615>, <214><710>, <194><806>, <171><901>]</trace>"}], "image": ["media/027102_00027102.jpg"]}
+{"id": "qa-trajectory_qa_28647", "conversations": [{"from": "human", "value": "<image-1>What is the required motion plan for the robotic arm to achieve \"retrieve the stuffed duck from the counter\"? Present trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<480><171>, <464><238>, <429><321>, <373><413>, <317><504>, <302><567>]</trace>"}], "image": ["media/028647_00028647.jpg"]}
+{"id": "qa-trajectory_qa_30042", "conversations": [{"from": "human", "value": "<image-1>To successfully execute \"move towards the basil bottle\", define the robotic arm's movement. Output trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<448><44>, <444><163>, <440><274>, <437><381>, <429><488>, <429><599>]</trace>"}], "image": ["media/030042_00030042.jpg"]}
+{"id": "qa-trajectory_qa_30783", "conversations": [{"from": "human", "value": "<image-1>The task is \"move the silver pot to the right burner\". How should the robotic arm navigate to this? Generate trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<91><587>, <163><607>, <226><611>, <286><607>, <341><579>, <377><544>]</trace>"}], "image": ["media/030783_00030783.jpg"]}
+{"id": "qa-trajectory_qa_31726", "conversations": [{"from": "human", "value": "<image-1>To achieve \"move the pepper to the lower left side of the table\", what sequence of movements should the robotic arm execute? Please. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<294><512>, <278><635>, <270><722>, <258><802>, <250><865>, <242><937>]</trace>"}], "image": ["media/031726_00031726.jpg"]}
+{"id": "qa-trajectory_qa_32728", "conversations": [{"from": "human", "value": "<image-1>Articulate the robotic arm's planned movement to achieve \"retrieve the white basket from its current position\". As for the output, list trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<480><278>, <504><421>, <540><532>, <587><635>, <643><734>, <683><837>]</trace>"}], "image": ["media/032728_00032728.jpg"]}
+{"id": "qa-trajectory_qa_34787", "conversations": [{"from": "human", "value": "<image-1>Detail the robotic arm's motion to achieve \"place the silver lid on top of the silver pot\". Present the solution as trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<480><278>, <480><329>, <480><377>, <480><425>, <480><476>, <480><528>]</trace>"}], "image": ["media/034787_00034787.jpg"]}
+{"id": "qa-trajectory_qa_35268", "conversations": [{"from": "human", "value": "<image-1>Generate trajectory points detailing how the robotic arm must move to achieve \"place the red object into the sink\". Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<512><167>, <591><234>, <643><302>, <687><373>, <714><440>, <750><516>]</trace>"}], "image": ["media/035268_00035268.jpg"]}
+{"id": "qa-trajectory_qa_41139", "conversations": [{"from": "human", "value": "<image-1>Detail the specific steps for the robotic arm's movement to complete \"move the blue plastic bottle near the chocolate bar\", and provide trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<409><290>, <390><369>, <370><433>, <357><484>, <341><532>, <321><389>]</trace>"}], "image": ["media/041139_00041139.jpg"]}
+{"id": "qa-trajectory_qa_46338", "conversations": [{"from": "human", "value": "<image-1>To achieve \"move towards the sponge on the table\", what sequence of movements should the robotic arm execute? Please provide trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<948><464>, <812><405>, <724><381>, <662><353>, <636><329>, <623><310>]</trace>"}], "image": ["media/046338_00046338.jpg"]}
+{"id": "qa-trajectory_qa_62463", "conversations": [{"from": "human", "value": "<image-1>What actions must the robotic arm take to achieve \"place the water bottle on the counter\"? Output trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<558><575>, <562><500>, <562><437>, <562><373>, <562><313>, <562><250>]</trace>"}], "image": ["media/062463_00062463.jpg"]}
+{"id": "qa-trajectory_qa_65404", "conversations": [{"from": "human", "value": "<image-1>How can the robotic arm be guided to achieve \"place the blue chip bag into the top drawer\"? List trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<669><306>, <675><413>, <659><512>, <630><607>, <594><694>, <568><786>]</trace>"}], "image": ["media/065404_00065404.jpg"]}
+{"id": "qa-trajectory_qa_75197", "conversations": [{"from": "human", "value": "<image-1>For achieving \"place the blue chip bag on the counter\", detail the robotic arm's course of action with an output of trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<562><802>, <601><794>, <649><782>, <701><774>, <756><766>, <825><758>]</trace>"}], "image": ["media/075197_00075197.jpg"]}
+{"id": "qa-trajectory_qa_89536", "conversations": [{"from": "human", "value": "<image-1>The task is \"move towards the top drawer\". How should the robotic arm navigate to this? Generate trajectory points. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<974><405>, <942><512>, <854><623>, <750><675>, <640><675>, <562><591>]</trace>"}], "image": ["media/089536_00089536.jpg"]}
+{"id": "qa-trajectory_qa_96394", "conversations": [{"from": "human", "value": "<image-1>To fill the white paper cup on the table instead of the conveyor belt, what specific trajectory must the robot arm guide the glass along, starting from its current pouring position?. Output trajectory in 6 key points."}, {"from": "gpt", "value": "<trace>[<332><325>, <261><357>, <291><425>, <332><385>, <382><377>, <393><429>]</trace>"}], "image": ["media/096394_00096394.jpg"]}
+{"id": "qa-trajectory_qa_96614", "conversations": [{"from": "human", "value": "<image-1>What trajectory should the arm's gripper follow to place the container it is holding onto the empty white shelf in the background, ensuring none of the remaining contents spill?. Output trajectory in 7 key points."}, {"from": "gpt", "value": "<trace>[<615><373>, <602><282>, <668><242>, <736><246>, <799><270>, <868><286>, <912><361>]</trace>"}], "image": ["media/096614_00096614.jpg"]}

fsd_level_1_2_3/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "fsd_level_1_2_3",
+  "modality": [
+    "image"
+  ],
+  "category": "embodied_video_vqa",
+  "total": 229124,
+  "sampled": 20
+}

fsd_level_1_2_3/sampled.jsonl

@@ -0,0 +1,20 @@
+{"conversations": [{"from": "human", "value": "<image-1>\nBriefly describe this scene and locate all visible objects using <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>."}, {"from": "gpt", "value": "A <ref>robotic arm</ref><box>[<118><1><465><761>]</box> is hovering above a <ref>green cutting board</ref><box>[<135><225><320><419>]</box> on a stove, next to colorful <ref>red chili pepper</ref><box>[<502><535><587><721>]</box> and <ref>orange pumpkin</ref><box>[<504><119><592><227>]</box> pieces, surrounded by various utensils and objects."}], "image": ["media/008536_im_11.jpg"]}
+{"conversations": [{"from": "human", "value": "What is shown in this image? Describe the scene and annotate each object with <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>.\n<image-1>"}, {"from": "gpt", "value": "A <ref>blue spatula</ref><box>[<209><132><446><713>]</box> is positioned to the left of a <ref>green oval object</ref><box>[<516><539><644><712>]</box>, with a <ref>yellow plate</ref><box>[<345><723><645><996>]</box> below them. A <ref>red toy cup</ref><box>[<470><163><574><337>]</box> is to the right of the plate."}], "image": ["media/014194_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "Describe the image concisely. Identify and locate key objects with bounding boxes in the format <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>.\n<image-1>"}, {"from": "gpt", "value": "A <ref>robotic arm</ref><box>[<2><99><657><745>]</box> is positioned next to a <ref>blue play sink</ref><box>[<230><287><708><885>]</box> containing various colored objects, with a <ref>pink toy cup</ref><box>[<699><451><801><561>]</box> located on the surface to the right."}], "image": ["media/033618_im_12.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nGive a short description of this image, grounding each object with <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>."}, {"from": "gpt", "value": "A <ref>robotic arm</ref><box>[<240><0><746><779>]</box> is positioned above a white tray containing scattered black objects. The arm is moving forward, preparing to sweep the objects into a pile."}], "image": ["media/058220_im_17.jpg"]}
+{"conversations": [{"from": "human", "value": "Briefly describe this scene and locate all visible objects using <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>.\n<image-1>"}, {"from": "gpt", "value": "A <ref>robotic arm</ref><box>[<341><3><737><987>]</box> is positioned above a <ref>white basket</ref><box>[<80><449><420><741>]</box> on the wooden surface, with a <ref>blue cloth</ref><box>[<469><768><766><999>]</box> underneath it and a grey machine behind."}], "image": ["media/073470_im_15.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nIdentify all notable objects in the image and describe the scene. Mark object locations using <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>."}, {"from": "gpt", "value": "A <ref>silver colander</ref><box>[<634><369><886><640>]</box> is placed on a wooden table, with a <ref>white soup can</ref><box>[<526><525><755><916>]</box> in front. A <ref>robotic arm</ref><box>[<474><1><752><375>]</box> is positioned above the table, ready to move objects."}], "image": ["media/091039_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "Describe the scene in the image. Identify and locate all key objects using <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>.\n<image-1>"}, {"from": "gpt", "value": "A <ref>red cylinder</ref><box>[<231><571><328><692>]</box> is on a black table flanked by scattered <ref>blue objects</ref><box>[<593><600><812><744>]</box>."}], "image": ["media/103210_start_frame.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nObserve this robotic scene and describe all visible objects with their locations using <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>."}, {"from": "gpt", "value": "A <ref>robot arm</ref><box>[<840><0><999><596>]</box> approaches a <ref>7up can</ref><box>[<344><262><437><461>]</box> on a <ref>table</ref><box>[<32><143><853><996>]</box> next to a <ref>blue can</ref><box>[<275><139><360><312>]</box> and a <ref>bottle</ref><box>[<396><0><443><62>]</box>."}], "image": ["media/104301_start_frame.jpg"]}
+{"conversations": [{"from": "human", "value": "Give a short description of this image, grounding each object with <ref>object name</ref><box>[<x1><y1><x2><y2>]</box>.\n<image-1>"}, {"from": "gpt", "value": "An <ref>orange</ref><box>[<487><734><552><820>]</box> in a bottom <ref>drawer</ref><box>[<124><545><913><950>]</box>; a <ref>bowl</ref><box>[<340><185><460><309>]</box> and <ref>sponge</ref><box>[<200><340><332><445>]</box> on the <ref>counter</ref><box>[<165><120><859><549>]</box>."}], "image": ["media/110933_start_frame.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nTell me about the items in your view and their relationships to each other. Answer the question with scene graph."}, {"from": "gpt", "value": "<ref>yellow knife</ref><box>[<327><91><753><349>]</box> appears <pred>left to</pred><box>[<327><91><753><349>]</box><box>[<817><448><997><773>]</box> <ref>silver pot</ref><box>[<817><448><997><773>]</box>. <ref>yellow knife</ref><box>[<327><91><753><349>]</box> is positioned <pred>behind</pred><box>[<327><91><753><349>]</box><box>[<406><755><778><963>]</box> <ref>blue spatula</ref><box>[<406><755><778><963>]</box>. <ref>silver pot</ref><box>[<817><448><997><773>]</box> is located <pred>right to</pred><box>[<817><448><997><773>]</box><box>[<406><755><778><963>]</box> <ref>blue spatula</ref><box>[<406><755><778><963>]</box>."}], "image": ["media/125388_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nDescribe the objects you see and their relative positions. Answer the question with scene graph."}, {"from": "gpt", "value": "<ref>robotic arm</ref><box>[<252><1><505><492>]</box> is situated <pred>to the left of</pred><box>[<252><1><505><492>]</box><box>[<554><303><816><577>]</box> <ref>orange cheese slice</ref><box>[<554><303><816><577>]</box>. <ref>robotic arm</ref><box>[<252><1><505><492>]</box> appears <pred>to the right of</pred><box>[<252><1><505><492>]</box><box>[<140><273><250><433>]</box> <ref>blue spatula</ref><box>[<140><273><250><433>]</box>. <ref>orange cheese slice</ref><box>[<554><303><816><577>]</box> appears <pred>to the right of</pred><box>[<554><303><816><577>]</box><box>[<341><292><527><496>]</box> <ref>silver round lid</ref><box>[<341><292><527><496>]</box>. <ref>orange cheese slice</ref><box>[<554><303><816><577>]</box> is situated <pred>right to</pred><box>[<554><303><816><577>]</box><box>[<140><273><250><433>]</box> <ref>blue spatula</ref><box>[<140><273><250><433>]</box>. <ref>silver round lid</ref><box>[<341><292><527><496>]</box> appears <pred>right to</pred><box>[<341><292><527><496>]</box><box>[<140><273><250><433>]</box> <ref>blue spatula</ref><box>[<140><273><250><433>]</box>."}], "image": ["media/131720_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nWhat items can you see and describe their positional relationship? Answer the question with scene graph."}, {"from": "gpt", "value": "<ref>blue napkin</ref><box>[<117><199><816><793>]</box> is situated <pred>left to</pred><box>[<117><199><816><793>]</box><box>[<622><0><777><125>]</box> <ref>red toy cup</ref><box>[<622><0><777><125>]</box>. <ref>red toy cup</ref><box>[<622><0><777><125>]</box> is located <pred>right to</pred><box>[<622><0><777><125>]</box><box>[<0><601><102><767>]</box> <ref>white mushroom</ref><box>[<0><601><102><767>]</box>."}], "image": ["media/133990_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nDescribe the objects you see and their relative positions. Answer the question with scene graph."}, {"from": "gpt", "value": "<ref>robotic arm</ref><box>[<238><1><516><517>]</box> appears <pred>to the left of</pred><box>[<238><1><516><517>]</box><box>[<520><0><618><205>]</box> <ref>colorful cups</ref><box>[<520><0><618><205>]</box>. <ref>metal pot</ref><box>[<493><385><771><741>]</box> is positioned <pred>in front of</pred><box>[<493><385><771><741>]</box><box>[<520><0><618><205>]</box> <ref>colorful cups</ref><box>[<520><0><618><205>]</box>."}], "image": ["media/137818_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nList the objects you see and describe their positional relationships. Answer the question with scene graph."}, {"from": "gpt", "value": "<ref>cooking surface</ref><box>[<0><643><998><996>]</box> stands <pred>below</pred><box>[<0><643><998><996>]</box><box>[<636><209><877><523>]</box> <ref>circular burners</ref><box>[<636><209><877><523>]</box>. <ref>cooking surface</ref><box>[<0><643><998><996>]</box> can be seen <pred>below</pred><box>[<0><643><998><996>]</box><box>[<357><736><632><997>]</box> <ref>robotic arm</ref><box>[<357><736><632><997>]</box>. <ref>circular burners</ref><box>[<636><209><877><523>]</box> is situated <pred>below</pred><box>[<636><209><877><523>]</box><box>[<357><736><632><997>]</box> <ref>robotic arm</ref><box>[<357><736><632><997>]</box>."}], "image": ["media/141759_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nList the objects you see and describe their positional relationships. Answer the question with scene graph."}, {"from": "gpt", "value": "<ref>robotic arm</ref><box>[<375><0><708><749>]</box> can be seen <pred>contains</pred><box>[<375><0><708><749>]</box><box>[<543><855><654><996>]</box> <ref>red toy cup</ref><box>[<543><855><654><996>]</box>. <ref>white cooking surface</ref><box>[<240><543><686><999>]</box> appears <pred>to the left of</pred><box>[<240><543><686><999>]</box><box>[<769><228><866><377>]</box> <ref>yellow shell-shaped object</ref><box>[<769><228><866><377>]</box>. <ref>white cooking surface</ref><box>[<240><543><686><999>]</box> stands <pred>contains</pred><box>[<240><543><686><999>]</box><box>[<543><855><654><996>]</box> <ref>red toy cup</ref><box>[<543><855><654><996>]</box>. <ref>yellow shell-shaped object</ref><box>[<769><228><866><377>]</box> is positioned <pred>right to</pred><box>[<769><228><866><377>]</box><box>[<543><855><654><996>]</box> <ref>red toy cup</ref><box>[<543><855><654><996>]</box>."}], "image": ["media/151169_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nList the objects you see and describe their positional relationships. Answer the question with scene graph."}, {"from": "gpt", "value": "<ref>table</ref><box>[<66><165><902><998>]</box> appears <pred>to the left of</pred><box>[<66><165><902><998>]</box><box>[<826><0><999><590>]</box> <ref>robot</ref><box>[<826><0><999><590>]</box>. <ref>table</ref><box>[<66><165><902><998>]</box> appears <pred>contains</pred><box>[<66><165><902><998>]</box><box>[<548><402><695><510>]</box> <ref>RxBar blueberry</ref><box>[<548><402><695><510>]</box>. <ref>robot</ref><box>[<826><0><999><590>]</box> is situated <pred>right to</pred><box>[<826><0><999><590>]</box><box>[<548><402><695><510>]</box> <ref>RxBar blueberry</ref><box>[<548><402><695><510>]</box>."}], "image": ["media/179919_start_frame.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nDoes green bell-shaped object appear to the left of yellow pineapple slice?"}, {"from": "gpt", "value": "Find the <ref>green bell-shaped object</ref><box>[<324><817><459><997>]</box> and <ref>yellow pineapple slice</ref><box>[<740><317><832><447>]</box> in the image. <ref>green bell-shaped object</ref><box>[<324><817><459><997>]</box> appears <pred>to the left of</pred><box>[<324><817><459><997>]</box><box>[<740><317><832><447>]</box> <ref>yellow pineapple slice</ref><box>[<740><317><832><447>]</box>.\nSo the answer is Yes, green bell-shaped object is positioned to the left of yellow pineapple slice."}], "image": ["media/196652_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nIs it correct to say that orange cheese slice is behind blue spatula from where we are looking?"}, {"from": "gpt", "value": "Find the <ref>orange cheese slice</ref><box>[<388><333><530><527>]</box> and <ref>blue spatula</ref><box>[<466><537><573><764>]</box> in the image. <ref>orange cheese slice</ref><box>[<388><333><530><527>]</box> appears <pred>behind</pred><box>[<388><333><530><527>]</box><box>[<466><537><573><764>]</box> <ref>blue spatula</ref><box>[<466><537><573><764>]</box>.\nSo the answer is Yes, orange cheese slice is situated behind blue spatula."}], "image": ["media/200346_im_0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image-1>\nFrom our point of view, is blue tray behind orange?"}, {"from": "gpt", "value": "Find the <ref>blue tray</ref><box>[<741><245><999><732>]</box> and <ref>orange</ref><box>[<485><296><572><374>]</box> in the image. <ref>blue tray</ref><box>[<741><245><999><732>]</box> can be seen <pred>to the right of</pred><box>[<741><245><999><732>]</box><box>[<485><296><572><374>]</box> <ref>orange</ref><box>[<485><296><572><374>]</box>.\nSo the answer is No, blue tray is not positioned behind orange."}], "image": ["media/209523_start_frame.jpg"]}
+{"conversations": [{"from": "human", "value": "What is the <ref>robot arm</ref><box>[<15><2><464><633>]</box> doing in the image?"}, {"from": "gpt", "value": "The <ref>robot arm</ref><box>[<15><2><464><633>]</box> is positioned <pred>to the left of</pred><box>[<15><2><464><633>]</box><box>[<488><678><568><735>]</box> the <ref>blue marker</ref><box>[<488><678><568><735>]</box>, which is placed in the <ref>wooden tray</ref><box>[<465><623><702><847>]</box> on the <ref>desk</ref><box>[<17><559><997><996>]</box>. It appears to be involved in the task of placing or organizing items on the desk."}, {"from": "human", "value": "Where is the <ref>blue marker</ref><box>[<488><678><568><735>]</box> located?"}, {"from": "gpt", "value": "The <ref>blue marker</ref><box>[<488><678><568><735>]</box> is located inside the <ref>wooden tray</ref><box>[<465><623><702><847>]</box>, which is on the <ref>desk</ref><box>[<17><559><997><996>]</box>. The tray is used to organize or hold items, and the marker is neatly placed within it."}, {"from": "human", "value": "What is the purpose of the <ref>wooden tray</ref><box>[<465><623><702><847>]</box> on the <ref>desk</ref><box>[<17><559><997><996>]</box>?"}, {"from": "gpt", "value": "The <ref>wooden tray</ref><box>[<465><623><702><847>]</box> on the <ref>desk</ref><box>[<17><559><997><996>]</box> serves as an organizational tool to hold and contain items like the <ref>blue marker</ref><box>[<488><678><568><735>]</box>. It helps keep the desk tidy by providing a designated space for small objects, making it easier to find and access them when needed."}, {"from": "human", "value": "What can you infer about the setting based on the objects present?"}, {"from": "gpt", "value": "The setting appears to be a workspace or laboratory environment. The presence of a <ref>robot arm</ref><box>[<15><2><464><633>]</box> suggests automation or robotic tasks are being performed, possibly for organizing or sorting items. The <ref>wooden tray</ref><box>[<465><623><702><847>]</box> and <ref>blue marker</ref><box>[<488><678><568><735>]</box> indicate that the space is used for tasks that require writing or labeling. The overall setup suggests a focus on efficiency and organization, typical of a professional or technical environment."}], "image": ["media/225548_end_frame.jpg"]}

idkb/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "idkb",
+  "modality": [
+    "image"
+  ],
+  "category": "drive_qa",
+  "total": 950407,
+  "sampled": 20
+}

omni3d_arkitscenes/sampled.jsonl

@@ -0,0 +1,20 @@
+{"id": "omni_nuscenes_31891.521_00001723_000860", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: cabinet, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<4><0><301><278>] 3d_box [-1.33, -1.45, 2.92] [1.82, 0.79, 0.44] [0.87, 0.17, 0.75, 0.93, 0.67, 0.03]</box>"}], "image": ["media/000860_31891.521_00001723.jpg"]}
+{"id": "omni_nuscenes_99487.667_00003104_002844", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: sofa, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><0><878><815>] 3d_box [-0.38, -0.33, 2.12] [2.10, 0.89, 1.06] [0.99, 0.38, 0.93, 0.76, 0.97, 0.33]</box>"}], "image": ["media/002844_99487.667_00003104.jpg"]}
+{"id": "omni_nuscenes_135829.840_00002930_012358", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: chair, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><364><912><999>] 3d_box [-0.15, 0.50, 1.06] [0.47, 1.01, 0.67] [0.75, 0.07, 0.43, 0.99, 0.79, 0.09]</box> <box>2d_box [<0><249><708><915>] 3d_box [-0.27, 0.21, 1.58] [0.49, 1.01, 0.61] [0.76, 0.07, 0.40, 0.99, 0.77, 0.08]</box> <box>2d_box [<0><177><612><710>] 3d_box [-0.35, -0.09, 2.05] [0.48, 1.01, 0.60] [0.76, 0.07, 0.38, 0.98, 0.76, 0.07]</box> <box>2d_box [<0><385><328><856>] 3d_box [-0.91, 0.32, 2.14] [0.47, 1.02, 0.63] [0.76, 0.93, 0.59, 0.99, 0.22, 0.92]</box>"}], "image": ["media/012358_135829.840_00002930.jpg"]}
+{"id": "omni_nuscenes_183792.508_00000157_020278", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: sofa, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><708><292><999>] 3d_box [-1.34, 1.06, 2.77] [0.86, 0.75, 1.11] [0.02, 0.35, 0.00, 0.52, 0.97, 0.33]</box> <box>2d_box [<345><669><563><857>] 3d_box [-0.34, 1.28, 5.55] [1.52, 0.59, 1.23] [0.48, 0.00, 0.54, 1.00, 0.48, 0.00]</box>"}], "image": ["media/020278_183792.508_00000157.jpg"]}
+{"id": "omni_nuscenes_35343.254_00001508_028157", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: chair, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><0><296><517>] 3d_box [-1.37, -0.75, 2.40] [0.86, 0.98, 1.03] [0.77, 0.08, 0.32, 0.97, 0.39, 0.01]</box> <box>2d_box [<529><0><696><215>] 3d_box [0.57, -1.72, 4.17] [0.51, 0.86, 0.48] [0.80, 0.10, 0.21, 0.91, 0.30, 0.04]</box> <box>2d_box [<393><0><561><210>] 3d_box [-0.11, -1.76, 4.26] [0.55, 0.86, 0.48] [0.82, 0.88, 0.19, 0.89, 0.26, 0.94]</box>"}], "image": ["media/028157_35343.254_00001508.jpg"]}
+{"id": "omni_nuscenes_39564.168_00006958_029083", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: cabinet, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<446><0><989><531>] 3d_box [0.42, -0.56, 1.87] [0.88, 0.93, 0.63] [0.98, 0.37, 0.92, 0.78, 0.99, 0.39]</box>"}], "image": ["media/029083_39564.168_00006958.jpg"]}
+{"id": "omni_nuscenes_133582.780_00002858_036170", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: cabinet, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><57><324><815>] 3d_box [-0.94, -0.17, 1.86] [0.67, 0.86, 0.65] [0.84, 0.86, 0.25, 0.93, 0.27, 0.94]</box> <box>2d_box [<0><0><285><495>] 3d_box [-1.15, -0.62, 2.52] [0.58, 0.87, 0.59] [0.80, 0.90, 0.58, 0.99, 0.54, 1.00]</box> <box>2d_box [<628><0><868><362>] 3d_box [0.82, -1.00, 2.94] [0.70, 0.85, 0.56] [0.80, 0.10, 0.42, 0.99, 0.46, 0.00]</box> <box>2d_box [<134><0><356><283>] 3d_box [-1.00, -1.24, 3.36] [0.35, 0.87, 0.54] [0.80, 0.90, 0.58, 0.99, 0.54, 1.00]</box> <box>2d_box [<560><0><668><262>] 3d_box [0.51, -1.38, 3.54] [0.31, 0.84, 0.60] [0.98, 0.37, 0.90, 0.81, 0.98, 0.36]</box> <box>2d_box [<309><0><453><252>] 3d_box [-0.50, -1.45, 3.70] [0.30, 0.84, 0.60] [0.98, 0.37, 0.90, 0.81, 0.98, 0.36]</box>"}], "image": ["media/036170_133582.780_00002858.jpg"]}
+{"id": "omni_nuscenes_15215.028_00000162_039365", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: chair, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<146><627><309><837>] 3d_box [-1.01, 1.16, 4.23] [0.40, 0.95, 0.53] [0.51, 1.00, 0.43, 1.00, 0.52, 1.00]</box>"}], "image": ["media/039365_15215.028_00000162.jpg"]}
+{"id": "omni_nuscenes_15059.740_00002438_039562", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: chair, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<786><551><897><674>] 3d_box [2.09, 0.88, 6.74] [0.43, 0.90, 0.48] [0.61, 0.01, 0.02, 0.65, 0.39, 0.01]</box>"}], "image": ["media/039562_15059.740_00002438.jpg"]}
+{"id": "omni_nuscenes_2951.355_00001177_041275", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: table, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<459><10><999><470>] 3d_box [0.61, -0.95, 2.71] [1.50, 0.94, 0.70] [0.80, 0.10, 0.29, 0.95, 0.36, 0.02]</box>"}], "image": ["media/041275_2951.355_00001177.jpg"]}
+{"id": "omni_nuscenes_14469.409_00002648_067013", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: sink, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<491><492><790><778>] 3d_box [0.26, 0.12, 1.42] [0.40, 0.19, 0.46] [0.75, 0.07, 0.95, 0.72, 0.73, 0.05]</box>"}], "image": ["media/067013_14469.409_00002648.jpg"]}
+{"id": "omni_nuscenes_178610.437_00000643_072569", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: table, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><458><542><862>] 3d_box [-0.29, 0.26, 1.29] [0.30, 0.41, 0.56] [0.90, 0.81, 0.57, 1.00, 0.60, 0.99]</box>"}], "image": ["media/072569_178610.437_00000643.jpg"]}
+{"id": "omni_nuscenes_482510.134_00001603_072918", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: table, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><571><177><999>] 3d_box [-1.43, 0.65, 2.65] [0.74, 0.66, 0.43] [0.64, 0.02, 0.82, 0.89, 0.57, 0.00]</box>"}], "image": ["media/072918_482510.134_00001603.jpg"]}
+{"id": "omni_nuscenes_18077.902_00002618_074798", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: table, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<352><26><999><999>] 3d_box [0.45, -0.07, 1.31] [1.10, 0.59, 0.64] [0.86, 0.85, 0.76, 0.92, 0.73, 0.94]</box>"}], "image": ["media/074798_18077.902_00002618.jpg"]}
+{"id": "omni_nuscenes_19580.424_00004719_074922", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: toilet, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<627><323><999><921>] 3d_box [0.48, 0.15, 1.48] [0.42, 0.74, 0.64] [0.96, 0.69, 0.95, 0.72, 0.95, 0.71]</box>"}], "image": ["media/074922_19580.424_00004719.jpg"]}
+{"id": "omni_nuscenes_11336.727_00011057_090402", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: shelves, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><349><120><547>] 3d_box [-2.05, -0.13, 3.16] [0.93, 0.42, 0.56] [0.42, 0.99, 0.92, 0.77, 0.41, 0.99]</box>"}], "image": ["media/090402_11336.727_00011057.jpg"]}
+{"id": "omni_nuscenes_19978.858_00000577_101125", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: sink, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<189><0><491><126>] 3d_box [-0.38, -0.85, 1.92] [0.52, 0.16, 0.45] [0.92, 0.77, 0.84, 0.87, 0.90, 0.79]</box>"}], "image": ["media/101125_19978.858_00000577.jpg"]}
+{"id": "omni_nuscenes_100734.074_00002180_107733", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: cabinet, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<164><81><573><327>] 3d_box [-0.58, -1.25, 4.32] [1.99, 0.59, 0.47] [0.86, 0.85, 0.90, 0.81, 0.79, 0.91]</box> <box>2d_box [<682><18><809><198>] 3d_box [1.55, -1.94, 5.31] [0.52, 0.73, 0.50] [0.85, 0.85, 0.89, 0.81, 0.78, 0.91]</box> <box>2d_box [<774><0><957><166>] 3d_box [2.77, -2.70, 6.29] [1.00, 1.10, 0.56] [0.84, 0.86, 0.88, 0.82, 0.77, 0.92]</box>"}], "image": ["media/107733_100734.074_00002180.jpg"]}
+{"id": "omni_nuscenes_100762.062_00003860_107785", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: chair, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<605><164><999><767>] 3d_box [0.84, -0.17, 2.29] [0.63, 0.94, 0.72] [0.72, 0.95, 0.83, 0.88, 0.67, 0.97]</box>"}], "image": ["media/107785_100762.062_00003860.jpg"]}
+{"id": "omni_nuscenes_177646.359_00000769_112160", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: sink, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<456><612><908><999>] 3d_box [0.24, 0.28, 1.08] [0.46, 0.14, 0.37] [0.82, 0.11, 0.95, 0.71, 0.80, 0.10]</box>"}], "image": ["media/112160_177646.359_00000769.jpg"]}

omni3d_hypersim/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "omni3d_hypersim",
+  "modality": [
+    "image"
+  ],
+  "category": "3D Detection",
+  "total": 275395,
+  "sampled": 20
+}

omni3d_hypersim/sampled.jsonl

@@ -0,0 +1,20 @@
+{"id": "omni_hypersim_frame.0030.tonemap_019784", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: lamp, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<125><31><437><368>] 3d_box [-0.86, -0.89, 3.88] [1.68, 0.68, 0.33] [0.48, 1.00, 0.71, 0.95, 0.47, 1.00]</box> <box>2d_box [<0><456><78><499>] 3d_box [-2.34, -0.08, 4.04] [0.72, 0.04, 0.04] [0.48, 1.00, 0.71, 0.95, 0.47, 1.00]</box> <box>2d_box [<122><475><171><478>] 3d_box [-1.97, -0.10, 4.85] [0.33, 0.01, 0.01] [0.48, 0.00, 0.29, 0.95, 0.53, 0.00]</box> <box>2d_box [<215><477><301><489>] 3d_box [-1.55, -0.08, 5.63] [0.72, 0.04, 0.09] [0.48, 0.00, 0.29, 0.95, 0.53, 0.00]</box>"}], "image": ["media/019784_frame.0030.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0041.tonemap_040475", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: lamp, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<688><494><771><633>] 3d_box [1.15, 0.24, 4.36] [0.20, 0.48, 0.19] [0.72, 0.05, 0.68, 0.97, 0.46, 0.00]</box> <box>2d_box [<592><0><631><35>] 3d_box [0.85, -2.80, 6.57] [0.08, 0.12, 0.08] [0.71, 0.05, 0.37, 0.98, 0.32, 0.03]</box> <box>2d_box [<669><17><685><43>] 3d_box [1.39, -2.76, 6.78] [0.08, 0.12, 0.08] [0.71, 0.05, 0.37, 0.98, 0.32, 0.03]</box> <box>2d_box [<731><38><747><63>] 3d_box [1.93, -2.72, 6.98] [0.08, 0.12, 0.08] [0.71, 0.05, 0.37, 0.98, 0.32, 0.03]</box>"}], "image": ["media/040475_frame.0041.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0017.tonemap_048184", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: counter, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<844><460><999><751>] 3d_box [2.27, 0.32, 4.04] [1.20, 0.79, 0.55] [0.51, 0.00, 0.58, 0.99, 0.54, 0.00]</box> <box>2d_box [<753><475><957><682>] 3d_box [2.08, 0.31, 5.23] [1.20, 0.79, 0.55] [0.51, 0.00, 0.58, 0.99, 0.54, 0.00]</box>"}], "image": ["media/048184_frame.0017.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0012.tonemap_049998", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: chair, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><507><697><999>] 3d_box [-0.12, 0.42, 0.83] [0.51, 0.79, 0.49] [0.49, 1.00, 0.03, 0.67, 0.52, 1.00]</box> <box>2d_box [<446><505><751><999>] 3d_box [0.24, 0.44, 1.92] [0.51, 0.79, 0.49] [0.49, 1.00, 0.03, 0.67, 0.52, 1.00]</box>"}], "image": ["media/049998_frame.0012.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0044.tonemap_072565", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: picture, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<443><64><478><120>] 3d_box [-0.32, -2.52, 7.15] [0.31, 0.29, 0.01] [0.73, 0.06, 0.91, 0.79, 0.74, 0.06]</box>"}], "image": ["media/072565_frame.0044.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0081.tonemap_081845", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: door, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<478><188><573><773>] 3d_box [0.17, -0.08, 5.52] [0.74, 2.65, 0.05] [0.52, 0.00, 0.48, 1.00, 0.40, 0.01]</box> <box>2d_box [<681><265><827><826>] 3d_box [1.65, 0.20, 5.71] [0.74, 2.65, 0.05] [0.52, 1.00, 0.39, 0.99, 0.59, 0.99]</box>"}], "image": ["media/081845_frame.0081.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0071.tonemap_090650", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: table, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<527><706><619><747>] 3d_box [0.72, 1.67, 8.49] [0.86, 0.25, 0.37] [0.01, 0.40, 0.00, 0.56, 0.99, 0.38]</box> <box>2d_box [<413><635><563><699>] 3d_box [-0.20, 2.10, 14.57] [2.41, 0.70, 1.10] [0.01, 0.60, 1.00, 0.56, 0.01, 0.62]</box> <box>2d_box [<436><642><517><676>] 3d_box [-0.55, 2.86, 20.73] [1.87, 0.55, 0.94] [0.01, 0.40, 0.00, 0.56, 0.99, 0.38]</box>"}], "image": ["media/090650_frame.0071.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0023.tonemap_113999", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: chair, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<355><616><486><846>] 3d_box [-0.67, 1.34, 6.98] [0.95, 1.21, 0.91] [0.34, 0.97, 0.99, 0.60, 0.33, 0.97]</box>"}], "image": ["media/113999_frame.0023.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0045.tonemap_122156", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: picture, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<886><18><999><267>] 3d_box [1.38, -0.80, 2.60] [0.33, 0.43, 0.01] [0.66, 0.97, 0.07, 0.75, 0.31, 0.96]</box>"}], "image": ["media/122156_frame.0045.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0092.tonemap_130996", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: window, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<307><0><497><434>] 3d_box [-0.75, -1.68, 6.87] [1.24, 2.31, 0.15] [0.53, 1.00, 0.89, 0.81, 0.57, 1.00]</box>"}], "image": ["media/130996_frame.0092.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0013.tonemap_134345", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: picture, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<18><326><44><354>] 3d_box [-2.64, -0.67, 4.87] [0.16, 0.10, 0.03] [0.61, 0.99, 0.58, 0.99, 0.55, 1.00]</box>"}], "image": ["media/134345_frame.0013.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0068.tonemap_138930", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: pillow, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><430><345><979>] 3d_box [-0.66, 0.18, 1.41] [0.59, 0.45, 0.51] [0.52, 0.00, 0.37, 0.98, 0.44, 0.00]</box> <box>2d_box [<102><462><425><774>] 3d_box [-0.53, 0.18, 2.08] [0.59, 0.42, 0.49] [0.51, 1.00, 0.62, 0.99, 0.56, 1.00]</box>"}], "image": ["media/138930_frame.0068.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0024.tonemap_157444", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: picture, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<639><595><663><676>] 3d_box [1.00, 0.67, 5.74] [0.28, 0.37, 0.02] [0.49, 1.00, 0.43, 1.00, 0.58, 0.99]</box> <box>2d_box [<634><396><683><546>] 3d_box [1.20, -0.16, 6.58] [0.63, 0.82, 0.05] [0.49, 1.00, 0.43, 1.00, 0.58, 0.99]</box> <box>2d_box [<0><249><27><331>] 3d_box [-4.15, -1.31, 7.12] [0.29, 0.30, 0.02] [0.45, 1.00, 1.00, 0.57, 0.53, 1.00]</box> <box>2d_box [<255><265><382><490>] 3d_box [-1.56, -0.78, 7.48] [0.88, 1.32, 0.04] [0.45, 1.00, 1.00, 0.57, 0.53, 1.00]</box>"}], "image": ["media/157444_frame.0024.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0057.tonemap_170232", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: mirror, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<201><374><324><653>] 3d_box [-1.78, 0.08, 6.49] [0.61, 1.49, 0.42] [0.50, 1.00, 0.89, 0.81, 0.56, 1.00]</box>"}], "image": ["media/170232_frame.0057.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0026.tonemap_186466", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: blinds, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<329><294><697><638>] 3d_box [0.15, -0.27, 9.58] [3.99, 2.79, 0.31] [0.16, 0.13, 0.00, 0.53, 0.84, 0.13]</box>"}], "image": ["media/186466_frame.0026.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0087.tonemap_192457", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: cabinet, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<232><443><561><576>] 3d_box [-1.02, 0.07, 9.09] [3.19, 0.96, 1.26] [0.41, 0.01, 0.00, 0.52, 0.59, 0.01]</box> <box>2d_box [<689><219><857><580>] 3d_box [2.73, -0.65, 9.10] [2.93, 2.39, 0.70] [0.50, 1.00, 0.48, 1.00, 0.50, 1.00]</box>"}], "image": ["media/192457_frame.0087.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0029.tonemap_209609", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: lamp, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<0><0><270><266>] 3d_box [-0.91, -0.91, 2.13] [0.81, 0.74, 0.28] [0.52, 0.00, 0.42, 0.99, 0.50, 0.00]</box>"}], "image": ["media/209609_frame.0029.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0016.tonemap_241366", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: shelves, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<887><504><948><712>] 3d_box [5.38, 1.00, 11.20] [1.75, 1.84, 0.33] [0.49, 0.00, 0.35, 0.98, 0.46, 0.00]</box> <box>2d_box [<871><511><921><687>] 3d_box [5.95, 1.08, 13.04] [1.75, 1.84, 0.33] [0.49, 0.00, 0.35, 0.98, 0.46, 0.00]</box> <box>2d_box [<726><481><765><547>] 3d_box [4.14, 0.18, 14.56] [0.56, 0.79, 0.19] [0.47, 1.00, 0.02, 0.65, 0.57, 1.00]</box>"}], "image": ["media/241366_frame.0016.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0094.tonemap_252986", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: pillow, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<587><418><642><464>] 3d_box [0.93, -0.36, 7.04] [0.35, 0.25, 0.27] [0.59, 0.99, 0.10, 0.80, 0.40, 0.99]</box> <box>2d_box [<561><415><628><461>] 3d_box [0.82, -0.42, 7.66] [1.04, 0.25, 0.17] [0.56, 1.00, 0.36, 0.98, 0.45, 1.00]</box>"}], "image": ["media/252986_frame.0094.tonemap.jpg"]}
+{"id": "omni_hypersim_frame.0098.tonemap_255997", "conversations": [{"from": "human", "value": "Detect all the objects in the image that belong to the category set: lamp, in 2D and 3D."}, {"from": "gpt", "value": "<box>2d_box [<671><356><863><767>] 3d_box [2.30, 0.42, 7.64] [1.28, 2.37, 1.12] [0.00, 0.43, 0.97, 0.68, 0.00, 0.49]</box>"}], "image": ["media/255997_frame.0098.tonemap.jpg"]}

oxe_grounding/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "oxe_grounding",
+  "modality": [
+    "video"
+  ],
+  "category": "grounding",
+  "total": 479688,
+  "sampled": 20
+}

oxe_grounding/sampled.jsonl

@@ -0,0 +1,20 @@
+{"image": ["media/013560_0000000728.jpg"], "conversations": [{"from": "human", "value": "Outline all regions that contain: The blue cup."}, {"from": "gpt", "value": "<ref>The blue cup</ref><box>[<892><54><1000><310>]</box>"}]}
+{"image": ["media/020898_0000000079.jpg"], "conversations": [{"from": "human", "value": "Draw bounding boxes around all: The robotic arm above the metal box."}, {"from": "gpt", "value": "<ref>The robotic arm above the metal box</ref><box>[<487><206><981><984>]</box>"}]}
+{"image": ["media/042174_0000000135.jpg"], "conversations": [{"from": "human", "value": "<image>\nDraw boxes around all: The red tomato.\nOutput format: <ref>label</ref><box>[<1><1><2><2>]</box>"}, {"from": "gpt", "value": "<ref>The red tomato</ref><box>[<428><325><501><430>]</box>"}]}
+{"image": ["media/047881_0000001456.jpg"], "conversations": [{"from": "human", "value": "Return bounding box coordinates for all: The green cup on the metal tray.\nCoordinates are [x1, y1, x2, y2], where (x1, y1) is top-left and (x2, y2) is bottom-right."}, {"from": "gpt", "value": "<ref>The green cup on the metal tray</ref><box>[<277><706><372><833>]</box>"}]}
+{"image": ["media/120022_0000001270.jpg"], "conversations": [{"from": "human", "value": "Draw bounding boxes around all instances of: The red and yellow condiment bottle hanging above the countertop."}, {"from": "gpt", "value": "<ref>red and yellow condiment bottle hanging above the countertop</ref><box>[<816><226><953><552>]</box>"}]}
+{"image": ["media/131264_0000001462.jpg"], "conversations": [{"from": "human", "value": "Locate all regions matching: The silver pot in the sink."}, {"from": "gpt", "value": "<ref>The silver pot in the sink</ref><box>[<625><441><815><686>]</box>"}]}
+{"image": ["media/168989_0000000494.jpg"], "conversations": [{"from": "human", "value": "Provide bounding boxes for every: The pineapple slices can."}, {"from": "gpt", "value": "<ref>The pineapple slices can</ref><box>[<734><134><875><334>]</box>"}]}
+{"image": ["media/226507_0000001793.jpg"], "conversations": [{"from": "human", "value": "Draw bounding boxes around all: The stove."}, {"from": "gpt", "value": "<ref>The stove</ref><box>[<672><174><1000><987>]</box>"}]}
+{"image": ["media/229511_0000000070.jpg"], "conversations": [{"from": "human", "value": "Outline all regions of: The orange toy mouse."}, {"from": "gpt", "value": "<ref>The orange toy mouse</ref><box>[<453><486><552><678>]</box>"}]}
+{"image": ["media/258890_0000002092.jpg"], "conversations": [{"from": "human", "value": "Outline all regions that contain: The blue dish rack."}, {"from": "gpt", "value": "<ref>The blue dish rack</ref><box>[<745><245><1125><620>]</box>"}]}
+{"image": ["media/262391_0000000419.jpg"], "conversations": [{"from": "human", "value": "Detect all instances of: The blue plastic spoon and return their bounding boxes."}, {"from": "gpt", "value": "<ref>The blue plastic spoon</ref><box>[<552><763><687><833>]</box>"}]}
+{"image": ["media/283105_0000000358.jpg"], "conversations": [{"from": "human", "value": "Provide bounding boxes for every: The pineapple juice can."}, {"from": "gpt", "value": "<ref>The pineapple juice can</ref><box>[<770><138><917><344>]</box>"}]}
+{"image": ["media/301569_0000002332.jpg"], "conversations": [{"from": "human", "value": "Find all bounding boxes for: The green corn toy next to the spoon."}, {"from": "gpt", "value": "<ref>green corn toy next to the spoon</ref><box>[<209><407><352><607>]</box>"}]}
+{"image": ["media/313543_0000000183.jpg"], "conversations": [{"from": "human", "value": "Locate all regions matching: The white plastic components."}, {"from": "gpt", "value": "<ref>The white plastic components</ref><box>[<441><401><816><750>]</box>"}]}
+{"image": ["media/316519_0000000858.jpg"], "conversations": [{"from": "human", "value": "Provide bounding boxes for every: The metal pan near the center of the wooden surface."}, {"from": "gpt", "value": "<ref>The metal pan near the center of the wooden surface</ref><box>[<406><281><681><502>]</box>"}]}
+{"image": ["media/353526_0000002276.jpg"], "conversations": [{"from": "human", "value": "Give the bounding box for each instance of: The black brush."}, {"from": "gpt", "value": "<ref>The black brush</ref><box>[<473><537><723><672>]</box>"}]}
+{"image": ["media/385107_0000002373.jpg"], "conversations": [{"from": "human", "value": "Draw boxes around all: The blue goggles inside the box."}, {"from": "gpt", "value": "<ref>The blue goggles inside the box</ref><box>[<240><581><354><684>]</box>"}]}
+{"image": ["media/438283_0000000086.jpg"], "conversations": [{"from": "human", "value": "Find all bounding boxes for: The metal bowl."}, {"from": "gpt", "value": "<ref>The metal bowl</ref><box>[<506><500><826><857>]</box>"}]}
+{"image": ["media/451158_0000000643.jpg"], "conversations": [{"from": "human", "value": "Detect all instances of: The orange clamps on the edges of the box and return their bounding boxes."}, {"from": "gpt", "value": "<ref>The orange clamps on the edges of the box</ref><box>[<96><198><178><341>]</box>"}]}
+{"image": ["media/466313_0000001785.jpg"], "conversations": [{"from": "human", "value": "Return bounding box coordinates for all: The pink plate in the sink."}, {"from": "gpt", "value": "<ref>The pink plate in the sink</ref><box>[<350><514><572><750>]</box>"}]}

pixmo_count_mc/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "pixmo_count_mc",
+  "modality": [
+    "image"
+  ],
+  "category": "counting",
+  "total": 34295,
+  "sampled": 20
+}

pixmo_count_number/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "pixmo_count_number",
+  "modality": [
+    "image"
+  ],
+  "category": "counting",
+  "total": 34295,
+  "sampled": 20
+}

pixmo_count_number/sampled.jsonl

@@ -0,0 +1,20 @@
+{"image": ["media/002764_b5513b2a5c2a9c80ad0a4c416c00f9be7a1d1f9ac1935d0efa0823c657c93581.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "5"}], "width": 1600, "height": 1067, "id": "pixmo_count_num_002764"}
+{"image": ["media/003413_57243ce23494ece12c04206073129545ce7f683008f310330feb3984af278919.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many cats are in the image?"}, {"from": "gpt", "value": "4"}], "width": 1000, "height": 750, "id": "pixmo_count_num_003413"}
+{"image": ["media/006157_70fb7ba4c3e64de571ed9e5916a0fcebf2f17d71326e1d24712c75ce92665b50.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "7"}], "width": 1600, "height": 1143, "id": "pixmo_count_num_006157"}
+{"image": ["media/008004_ec07b3b8f39278334d63b5423d7e6efa83a24752997570c817abf040cf34c70b.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are there in this image?"}, {"from": "gpt", "value": "3"}], "width": 1600, "height": 1598, "id": "pixmo_count_num_008004"}
+{"image": ["media/010394_4ce752bc49f31174b02aa2ea16182d4ea445475d0f2d7145bc40269f042159cc.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "7"}], "width": 1600, "height": 1067, "id": "pixmo_count_num_010394"}
+{"image": ["media/013023_71593e83bbe49035e49e578d023f43c712d85d3c53bc18f0c7b7e3faac650e2f.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many horses are in the image?"}, {"from": "gpt", "value": "1"}], "width": 1426, "height": 1000, "id": "pixmo_count_num_013023"}
+{"image": ["media/013053_8a72644b19691142faba84c7cf656b575a4766c33d8535e25a699cfc392ee438.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many guitars are there in this image?"}, {"from": "gpt", "value": "0"}], "width": 1600, "height": 1200, "id": "pixmo_count_num_013053"}
+{"image": ["media/013660_23c3f6ebf2f8796a031bf0cbef1193fb7ad547bee34cda3cbee9987bad9eae09.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "3"}], "width": 1223, "height": 1599, "id": "pixmo_count_num_013660"}
+{"image": ["media/014272_3a2d83d51c6687a43f82c1c5a36455e3fb0d46e0142349e36a718233d7b0ebdc.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many books are in the image?"}, {"from": "gpt", "value": "1"}], "width": 1600, "height": 1279, "id": "pixmo_count_num_014272"}
+{"image": ["media/016601_7a3df83eab0eefdec6a7fc41836ba1b9f2ef815d6723c58b5e09e5eb2ff750fa.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "1"}], "width": 2048, "height": 1291, "id": "pixmo_count_num_016601"}
+{"image": ["media/019583_49c6604341bf1ebfcf8c2bf93bf7251beea5ada49fb92ab088d3b5e065415db9.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many guitars are in the image?"}, {"from": "gpt", "value": "4"}], "width": 800, "height": 600, "id": "pixmo_count_num_019583"}
+{"image": ["media/020454_54cbfd8f201043a8173ba8aed042cfa2659222e5ddf5115c628e62e6e5dc9783.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "3"}], "width": 1600, "height": 1063, "id": "pixmo_count_num_020454"}
+{"image": ["media/021172_6831c2dc87e1793c8dc825c1155cb7b9f8735e336ca5ea0b3b7454ad2a701399.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "2"}], "width": 1600, "height": 1593, "id": "pixmo_count_num_021172"}
+{"image": ["media/023620_d74a5fb3a01bddba01dc74d411493fa5053c84a1d48e05ea699e9ab837020f77.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "3"}], "width": 1599, "height": 1068, "id": "pixmo_count_num_023620"}
+{"image": ["media/025501_d4bcf29ff4a1e98db848d85bdaa97e960eeb19277d56ccec91a5640882b91216.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people can you see in the image?"}, {"from": "gpt", "value": "6"}], "width": 1024, "height": 512, "id": "pixmo_count_num_025501"}
+{"image": ["media/030201_f97b4528c65655838d32ddffa0ff1adb2ce964a954d85fa0d04fa76a3bd559b1.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "1"}], "width": 1600, "height": 1067, "id": "pixmo_count_num_030201"}
+{"image": ["media/030987_69b15314aa324050b04810094593d6d3890e7f4e75079d94138198b76b0aa786.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are in the image?"}, {"from": "gpt", "value": "1"}], "width": 1600, "height": 1200, "id": "pixmo_count_num_030987"}
+{"image": ["media/031537_ab963e33b1243a1fde80897c0fd0ba9c8fc9b1f1fff4de9bf57f6448ce4cfde0.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many people are there in this image?"}, {"from": "gpt", "value": "1"}], "width": 2047, "height": 1299, "id": "pixmo_count_num_031537"}
+{"image": ["media/034139_3f99ba6bb8763172ecaffadd1481ae5440906f93541fe794ced46b624f209a58.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nCount the number of people in the image."}, {"from": "gpt", "value": "3"}], "width": 1599, "height": 1084, "id": "pixmo_count_num_034139"}
+{"image": ["media/034153_08e0c82fbc91c044e02a2d8a93b53b666fdf81cc0e8a694bc4aa50f0961d9bf4.jpg"], "task": "Count", "conversations": [{"from": "human", "value": "<image>\nHow many horses are there in this image?"}, {"from": "gpt", "value": "2"}], "width": 1155, "height": 1600, "id": "pixmo_count_num_034153"}

refspatial_2d_reasoning_template_qa/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "refspatial_2d_reasoning_template_qa",
+  "modality": [
+    "image"
+  ],
+  "category": "spatial_understanding",
+  "total": 460615,
+  "sampled": 20
+}

refspatial_simulator_metadata/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "refspatial_simulator_metadata",
+  "modality": [
+    "image"
+  ],
+  "category": "spatial_understanding",
+  "total": 245926,
+  "sampled": 20
+}

refspatial_simulator_metadata/sampled.jsonl

@@ -0,0 +1,20 @@
+{"conversations": [{"from": "human", "value": "<image>\nPoint to the the tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<652><43>]"}, {"from": "human", "value": "In the image, there is a the second object from left to right. Pinpoint a point to the the second object from left to right. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<453><348>]"}, {"from": "human", "value": "Give me the position of the right watch. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<574><480>]"}, {"from": "human", "value": "Select a location on the the shorter watch. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<344><537>]"}, {"from": "human", "value": "Please point out the the leftmost object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<344><537>]"}, {"from": "human", "value": "Please point out the the nearest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<669><783>]"}, {"from": "human", "value": "Please point to the the remote control. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<669><783>]"}, {"from": "human", "value": "Find a few locations in the blank space located 0.130m on the left of the table lamp. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<514><383>]"}, {"from": "human", "value": "Find several locations within the vacant space located on the left of the beer can. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<385><407>]"}, {"from": "human", "value": "Give me the position of a free location located on the face of the taller watch. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<439><546>]"}], "image": ["media/023627_4f70bf29c905af26.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nCan you point to a free location that satisfies the following two spatial constraints: 1. on the front side of the rightmost gray igniter; 2. at the face of the hamster? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<692><688>]"}, {"from": "human", "value": "Point to the vacant area in the image that satisfies the following two spatial constraints: 1. on the headband and ear cups of the headset; 2. on the back side of the second nearest igniter. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<340><703>]"}, {"from": "human", "value": "Find several locations within the vacant space that satisfies the following two spatial constraints: 1. at the face of the nearest red igniter; 2. on the left side of the brown hamster. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<443><703>]"}, {"from": "human", "value": "Point to the vacant area in the image that satisfies the following two spatial constraints: 1. on the headband and ear cups of the biggest gray igniter; 2. on the right side of the headset. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<375><494>]"}, {"from": "human", "value": "Locate the blank area in the image that satisfies the following two spatial constraints: 1. on the front and back of the vodka; 2. on the headband and ear cups of the third farthest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<607><426>]"}], "image": ["media/034448_28910de1f1736866.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nIndicate several points within the vacant area that lies between the seventh object from left to right and the fourth igniter from right to left. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<568><441>]"}, {"from": "human", "value": "Choose a point between the pear and the nearest gray igniter in the vacant region. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<336><717>]"}, {"from": "human", "value": "Identify several spots within the vacant space that's between the biggest gray igniter and the green pear. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<419><502>]"}, {"from": "human", "value": "What is the location of the object that stands as the third closest to the smallest red igniter? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<550><798>]"}, {"from": "human", "value": "Identify which object is located as the second nearest to the leftmost igniter. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<494><500>]"}, {"from": "human", "value": "Choose which igniter is the fourth nearest neighbor to the leftmost gray igniter. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<527><317>]"}, {"from": "human", "value": "Point to the third closest igniter to the leftmost object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<392><313>]"}, {"from": "human", "value": "Identify which red igniter is the closest to the shortest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<724><563>]"}, {"from": "human", "value": "Point out the red igniter that is the closest to the fourth farthest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<724><563>]"}, {"from": "human", "value": "Point out the second closest gray igniter to the leftmost gray igniter. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<527><317>]"}], "image": ["media/034644_eb4c4719d91c74f9.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nPoint to a point in vacant space on the desktop that is between the leftmost object and the duckling and is positioned on the face of the teddy bear. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<561><572>]"}, {"from": "human", "value": "Point out the free space on the desktop that is between the leftmost object and the duckling and is positioned at the barrel of the pistol. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<496><577>]"}], "image": ["media/043134_0d2e983fdb89e435.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nLocate a free spot on the desktop that is between the apple and the flashlight and is positioned on the peel of the second farthest object from the top. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<558><659>]"}, {"from": "human", "value": "Point to a free spot on the desktop that is between the tallest object and the flashlight and is positioned at the face of the duckling. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<577><656>]"}], "image": ["media/046042_fe3e8c159dc9fbb9.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nLocate a free spot on the desktop that is between the shortest object and the chinaware and is positioned on the side where the third object from left to right emits light. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<540><624>]"}, {"from": "human", "value": "Locate a point within vacant space on the desktop that is between the checkbook and the third tallest object and is positioned on the side of the second object from left to right with the handle loop. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<547><624>]"}], "image": ["media/052637_236044f5f0b72a76.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nSelect a location on the the tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<474><396>]"}, {"from": "human", "value": "Please point to the the phonograph record. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<267><359>]"}, {"from": "human", "value": "In the image, there is a the drone. Pinpoint a point to the the drone. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<701><502>]"}, {"from": "human", "value": "Locate a point on the the rightmost object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<701><502>]"}, {"from": "human", "value": "In the image, there is a the fourth tallest object. Pinpoint a point to the the fourth tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<271><580>]"}, {"from": "human", "value": "Choose the point of the goggles. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<490><604>]"}, {"from": "human", "value": "Please point to the the bagel. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<367><648>]"}, {"from": "human", "value": "Can you point to a free location located 0.320m on the front of the flower arrangement? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<455><673>]"}, {"from": "human", "value": "Locate several points within vacant space located from the horn of the second tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<223><673>]"}, {"from": "human", "value": "Locate several spots within the vacant area situated 0.140m on the left of the rightmost object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<558><587>]"}], "image": ["media/071441_b313b1a79250b4c7.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nPoint to the the detergent. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<509><283>]"}, {"from": "human", "value": "Choose the point of the coffeepot. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<427><233>]"}, {"from": "human", "value": "Choose the point of the birthday cake. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<498><378>]"}, {"from": "human", "value": "Select a point on the the microphone. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<477><541>]"}, {"from": "human", "value": "Give me the position of the fifth farthest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<648><617>]"}, {"from": "human", "value": "Choose the point of the rightmost object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<648><617>]"}, {"from": "human", "value": "In the image, there is a the left zucchini. Pinpoint a point to the the left zucchini. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<381><774>]"}, {"from": "human", "value": "Choose the point of the shorter zucchini. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<381><774>]"}, {"from": "human", "value": "Locate several spots in free space located on the front of the detergent. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<511><348>]"}, {"from": "human", "value": "Pinpoint the vacant space located on the front of the coffeepot. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<417><403>]"}], "image": ["media/077361_7ad60beb0909a4f5.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nIndicate several points within the vacant area that lies between the fifth object from left to right and the pancake. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<553><671>]"}, {"from": "human", "value": "Identify which object is located as the second nearest to the pancake. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<259><731>]"}, {"from": "human", "value": "What is the location of the object that stands as the fifth closest to the third tallest object? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<666><215>]"}, {"from": "human", "value": "Where can I find a free point that satisfies the following two spatial constraints: 1. on the spine side of the Bible; 2. at the stem of the shortest object? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<671><780>]"}, {"from": "human", "value": "Identify some points within the free space that satisfies the following two spatial constraints: 1. on the back side of the second tallest object; 2. on the right side of the icecream. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<355><318>]"}, {"from": "human", "value": "What is the location of a free spot that satisfies the following two spatial constraints: 1. on the spine side of the Bible; 2. at the stem of the shortest object? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<671><780>]"}], "image": ["media/085647_2138177a3bdeb4c4.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nPoint to a point within vacant space on the desktop that is between the wallet and the rightmost object and is positioned at the tip of the tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<499><721>]"}, {"from": "human", "value": "Point out the free space on the desktop that is between the leftmost object and the shortest object and is positioned on the front-facing side of the third shortest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<530><720>]"}], "image": ["media/095907_86ecb97f346e6884.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nCan you find a vacant point located between the shortest object and the nearer muffin? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<359><661>]"}, {"from": "human", "value": "Give me the position of a free point between the shortest object and the third object from right to left. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<359><661>]"}, {"from": "human", "value": "Locate some points within the free area in between the fire extinguisher and the fourth tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<494><549>]"}, {"from": "human", "value": "Locate which object ranks third in proximity to the fourth tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<234><770>]"}, {"from": "human", "value": "Locate which object is the third closest to the platter. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<234><770>]"}, {"from": "human", "value": "Locate several points within the vacant space that satisfies the following two spatial constraints: 1. on the nozzle side of the farthest object; 2. on the right side of the farther muffin. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<354><457>]"}], "image": ["media/108921_fc706d3906f767b2.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nLocate several spots within the vacant area that satisfies the following two spatial constraints: 1. on the back side of the shortest chalice; 2. on the right side of the gourd. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<498><388>]"}, {"from": "human", "value": "Point to any free space in the image that satisfies the following two spatial constraints: 1. on the back side of the videotape; 2. on the right side of the third farthest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<347><588>]"}, {"from": "human", "value": "Locate several points within the vacant space that satisfies the following two spatial constraints: 1. on the back side of the shortest chalice; 2. on the left side of the second nearest chalice. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<497><588>]"}], "image": ["media/131542_a665f0702d31cb78.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nPoint to a point in free space on the desktop that is between the nearer Lego and the sunglasses and is positioned on the front-facing side of the second object from left to right. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<476><530>]"}, {"from": "human", "value": "Locate a point within vacant space on the desktop that is between the leftmost object and the shortest object and is positioned on the front-facing side of the right figurine. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<531><528>]"}, {"from": "human", "value": "Point to a vacant spot on the desktop that is between the nearer Lego and the sunglasses and is positioned holding a crossbow in the hands of the farthest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<521><528>]"}], "image": ["media/139159_b69d46398be46da0.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nLocate a point on the alcohol that is at the left edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<309><226>]"}, {"from": "human", "value": "Point to the teddy bear at the right edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<799><659>]"}, {"from": "human", "value": "Point out the vodka located in the right edge section of the desktop. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<672><350>]"}, {"from": "human", "value": "Point out the object located in the far left corner section of the desktop. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<309><226>]"}, {"from": "human", "value": "Point out the object situated at the near left corner. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<315><687>]"}, {"from": "human", "value": "Locate the object that is at the near right corner. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<799><659>]"}, {"from": "human", "value": "In the image, there is a desktop. Pinpoint the vacant space at the far edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<554><308>]"}, {"from": "human", "value": "In the image, there is a desktop. Pinpoint the vacant space at the near edge of the desktop. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<467><668>]"}, {"from": "human", "value": "Locate several points within the vacant space at the left edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<367><356>]"}, {"from": "human", "value": "Locate several points within the vacant space at the right edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<702><416>]"}], "image": ["media/141402_bd16f93000ad2e72.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nLocate several points within the vacant space that satisfies the following two spatial constraints: 1. on the back side of the second shortest object; 2. on the face of the sunflower. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<597><490>]"}, {"from": "human", "value": "Pinpoint some points within the vacant space that satisfies the following two spatial constraints: 1. on the face of the fifth tallest object; 2. on the right side of the sunflower. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<288><613>]"}, {"from": "human", "value": "Can you find a vacant point that satisfies the following two spatial constraints: 1. on the back side of the sixth tallest object; 2. on the face of the sunflower? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<597><490>]"}, {"from": "human", "value": "Point to any free space in the image that satisfies the following two spatial constraints: 1. on the face of the chalice; 2. on the right side of the sunflower. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<312><510>]"}, {"from": "human", "value": "Locate some places in the vacant space that satisfies the following two spatial constraints: 1. on the face of the sunflower; 2. on the left side of the beanbag. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<288><613>]"}, {"from": "human", "value": "Point to any free space in the image that satisfies the following two spatial constraints: 1. on the face of the fifth tallest object; 2. on the left side of the sunflower. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<288><613>]"}], "image": ["media/195947_75bdeb3758c84dcd.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nPoint to the vacant area between the nearest object and the fourth nearest object in the image. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<460><662>]"}, {"from": "human", "value": "Locate any unoccupied position between the second nearest object and the taller microphone in the image. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<617><533>]"}, {"from": "human", "value": "Indicate several points within the vacant area that lies between the shortest object and the third farthest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<460><662>]"}, {"from": "human", "value": "Point to a free spot between the nearer microphone and the farther microphone. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<461><486>]"}, {"from": "human", "value": "Pinpoint some spots within the free space that is in between the pop soda and the headset. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<565><554>]"}, {"from": "human", "value": "Point to any empty space between the farther microphone and the shortest object in the image. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<430><597>]"}, {"from": "human", "value": "Point to any empty space between the shortest object and the shoulder bag in the image. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<460><662>]"}, {"from": "human", "value": "Point out which object is positioned as the nearest to the shorter microphone. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<305><848>]"}, {"from": "human", "value": "Identify which object is the fourth nearest to the headset. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<305><848>]"}, {"from": "human", "value": "Locate several points within the vacant space that satisfies the following two spatial constraints: 1. on the front side of the taller microphone; 2. on the headband of the headset. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<518><358>]"}], "image": ["media/206179_12530d9e33b98414.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nPlease point out the the bowl. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<550><241>]"}, {"from": "human", "value": "Locate a point on the the leftmost object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<439><291>]"}, {"from": "human", "value": "What is the 2D location of the mug? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<571><285>]"}, {"from": "human", "value": "Please point out the the Lego. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<467><694>]"}, {"from": "human", "value": "Give me the position of the nearest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<467><694>]"}, {"from": "human", "value": "Find several locations within the vacant space located 0.200m on the left of the bowl. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<490><265>]"}, {"from": "human", "value": "Locate several points in vacant space located 0.320m on the front of the chalice. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<432><443>]"}, {"from": "human", "value": "Identify a few places in vacant region located on the side of the second shortest object with the handle. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<582><443>]"}, {"from": "human", "value": "Identify a few spots in free region located on the front-facing side of the nearest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<463><827>]"}, {"from": "human", "value": "Locate the object that is at the far edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<550><241>]"}], "image": ["media/217703_0cc7047723758461.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nIn the image, there is a computer monitor at the left edge. Pinpoint a point to the computer monitor at the left edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<392><230>]"}, {"from": "human", "value": "Identify the bottle located at the left edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<259><524>]"}, {"from": "human", "value": "Point to the bottle at the right edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<610><281>]"}, {"from": "human", "value": "Locate a point on the teddy bear that is at the right edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<645><472>]"}, {"from": "human", "value": "In the image, there is a steak at the right edge. Pinpoint a point to the steak at the right edge. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<688><756>]"}, {"from": "human", "value": "Point to the object at the far left corner. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<392><230>]"}, {"from": "human", "value": "Point to the object at the far right corner. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<610><281>]"}, {"from": "human", "value": "In the image, there is a object at the near right corner. Pinpoint a point to the object at the near right corner. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<688><756>]"}, {"from": "human", "value": "Identify a few spots in the free region at the far edge of the desktop. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<492><366>]"}, {"from": "human", "value": "Locate any vacant position at the left edge of the desktop in the image. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<279><756>]"}], "image": ["media/229758_a7aba2470376cf3d.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nPlease point out the the sixth object from right to left. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<365><630>]"}, {"from": "human", "value": "Identify a spot on the the nearest white shampoo. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<615><646>]"}, {"from": "human", "value": "Point out the the second white shampoo from right to left. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<615><646>]"}, {"from": "human", "value": "What is the 2D location of the rightmost object? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<814><791>]"}, {"from": "human", "value": "Locate the the blue vodka. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<814><791>]"}, {"from": "human", "value": "Point to the vacant area located 0.160m on the front label of the rightmost blue shampoo in the image. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<501><368>]"}, {"from": "human", "value": "Identify a few places in the vacant region located on the front label of the rightmost blue shampoo. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<445><368>]"}, {"from": "human", "value": "Identify a few places in the vacant region located 0.280m on the front label of the rightmost blue shampoo. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<456><368>]"}, {"from": "human", "value": "Locate several points within the vacant space positioned on the front-facing side of the farthest white shampoo. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<460><371>]"}, {"from": "human", "value": "Point to a free spot located 0.290m on the front label of the fourth shampoo from right to left. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<616><448>]"}], "image": ["media/238453_ef8e629b533294a6.jpg"]}
+{"conversations": [{"from": "human", "value": "<image>\nChoose the point of the left chocolate muffin. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<494><624>]"}, {"from": "human", "value": "Select a location on the the bigger chocolate muffin. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<494><624>]"}, {"from": "human", "value": "Select a location on the the seventh farthest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<435><781>]"}, {"from": "human", "value": "What is the 2D location of the nearest beige muffin? Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<435><781>]"}, {"from": "human", "value": "Please point out the the smaller chocolate muffin. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<580><944>]"}, {"from": "human", "value": "Locate a point on the the nearest muffin. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<580><944>]"}, {"from": "human", "value": "Select a few points in free space located on the label of the tallest object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<520><291>]"}, {"from": "human", "value": "Identify some points in free space located on the left of the gray chalice. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<514><314>]"}, {"from": "human", "value": "Find a few locations in blank space located 0.160m with the flap open on the leftmost object. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<465><426>]"}, {"from": "human", "value": "Point to a free spot located 0.310m on the front of the second beige muffin from left to right. Your answer should be formatted as a list of tuples, i.e. [(x1, y1)], where each tuple contains the x and y coordinates of a point satisfying the conditions above."}, {"from": "gpt", "value": "[<524><525>]"}], "image": ["media/239465_a84a04fa631f7789.jpg"]}

sharerobot/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "sharerobot",
+  "modality": [
+    "video"
+  ],
+  "category": "embodied_video_vqa",
+  "total": 1025900,
+  "sampled": 20
+}

sharerobot/sampled.jsonl

@@ -0,0 +1,20 @@
+{"video": ["media/002689_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nIs this \"approach the wipe roll\" you're focused on right now?"}, {"from": "gpt", "value": "no"}]}
+{"video": ["media/073201_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nIs this \"lower and sweep the objects across the surface\" you're focused on right now?"}, {"from": "gpt", "value": "no"}]}
+{"video": ["media/280150_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nBased on the current situation, what is expected to happen after \"move the banana towards the bowl\"?"}, {"from": "gpt", "value": "place the banana in the bowl"}]}
+{"video": ["media/289666_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nAfter \"move the spray bottle\", what's the most probable next event?"}, {"from": "gpt", "value": "place the spray bottle down"}]}
+{"video": ["media/296578_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nConsidering the current sequence of tasks, what's likely to occur after \"move towards the red object\"?"}, {"from": "gpt", "value": "grab the red object"}]}
+{"video": ["media/313058_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nGiven the current situation, what action can be taken?"}, {"from": "gpt", "value": "place the spatula down on the table"}]}
+{"video": ["media/356385_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nConsidering the circumstances, what task can you begin now?"}, {"from": "gpt", "value": "lift the spoon"}]}
+{"video": ["media/435098_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nWhat task did you just finish?"}, {"from": "gpt", "value": "move towards the cylinder"}]}
+{"video": ["media/532401_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nThe goal is \"placing a piece of fruit into a bowl\". After completing 1-\"reach for the fruit\", 2-\"grab the fruit\", 3-\"lift the fruit\", what are the next five steps you should take?"}, {"from": "gpt", "value": "4-move the fruit towards the bowl, 5-place the fruit inside the bowl, 6-release the fruit"}]}
+{"video": ["media/608093_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nHere's what's been done so far: 1-\"reach for the double-square object\", 2-\"pick up the double-square object\", 3-\"move the double-square object towards the blue board\". What are the next five tasks to take toward the goal of \"inserting a double-square object into its matching slot\"?"}, {"from": "gpt", "value": "4-insert the double-square object into the matching slot, 5-release the double-square object, 6-retract from the board"}]}
+{"video": ["media/615448_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nGiven the progress so far: 1-\"pick up the brown cylinder\", what's the next set of five steps to move closer to \"inserting a brown cylinder into a designated slot\"?"}, {"from": "gpt", "value": "2-move the cylinder towards the slot, 3-insert the cylinder into the slot, 4-release the cylinder"}]}
+{"video": ["media/691998_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nGiven the goal of \"moving a cloth to the left side of a table\", what is the most logical next move?"}, {"from": "gpt", "value": "drag the cloth to the left"}]}
+{"video": ["media/761319_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nGiven your progress so far (1-\"pick up the tea bag\"), what's the next step toward achieving \"making tea\"?"}, {"from": "gpt", "value": "place the tea bag in the cup"}]}
+{"video": ["media/810208_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nConsidering the goal of \"picking up a red object from a table\", and having done 1-\"move towards the red object\", 2-\"grab the red object\", what should you do next?"}, {"from": "gpt", "value": "lift the red object"}]}
+{"video": ["media/825981_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nHas \"place the can on the near edge of the table\" been fully carried out?"}, {"from": "gpt", "value": "no"}]}
+{"video": ["media/828381_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nCan we say that \"lift the fork\" was accomplished?"}, {"from": "gpt", "value": "no"}]}
+{"video": ["media/860408_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nWas \"place the clothes inside the washing machine\" finalized?"}, {"from": "gpt", "value": "no"}]}
+{"video": ["media/878157_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nHas \"place the bowl onto the board\" been fully carried out?"}, {"from": "gpt", "value": "no"}]}
+{"video": ["media/895916_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nWas \"place the banana in the mug\" finalized?"}, {"from": "gpt", "value": "no"}]}
+{"video": ["media/979630_video.mp4"], "conversations": [{"from": "human", "value": "<video>\nWas \"lift the towel\" finalized?"}, {"from": "gpt", "value": "yes"}]}

spar_scannet_train_depth_prediction_oc_fill_105066/meta.json

@@ -0,0 +1,9 @@
+{
+  "name": "spar_scannet_train_depth_prediction_oc_fill_105066",
+  "modality": [
+    "image"
+  ],
+  "category": "spatial_understanding",
+  "total": 105066,
+  "sampled": 20
+}