file_name stringlengths 31 31 | im_path stringlengths 46 46 | caption stringlengths 2 104 | bbox stringlengths 21 45 | prompt stringlengths 2.03k 2.52k | texts stringlengths 2 61 | sentence stringlengths 42 524 | outp_org listlengths 4 4 | formatted_sentence stringlengths 37 71 | modified_sentence stringlengths 38 518 | target listlengths 4 4 |
|---|---|---|---|---|---|---|---|---|---|---|
COCO_train2014_000000570786.jpg | refcoco_images/COCO_train2014_000000570786.jpg | second man from right | [323.51,,80.29,418.97,298.79] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Man'] | Detected Man at location [196, 114, 361, 355]
Detected second man from right at location [321, 82, 418, 299]
| [
321,
82,
418,
299
] | Detected man at location [321, 82, 418, 299] | Detected Man at location [196, 114, 361, 355]
Detected man at location [321, 82, 418, 299] | [
321,
82,
418,
299
] |
COCO_train2014_000000523466.jpg | refcoco_images/COCO_train2014_000000523466.jpg | sandwich left most | [,,0.,,,120.63,218.29,293.92] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Sandwich'] | Detected Sandwich at location [4, 116, 633, 392]
Detected Sandwich at location [2, 119, 402, 364]
Detected Sandwich at location [52, 180, 402, 363]
Detected sandwich left most at location [0, 122, 219, 294]
| [
0,
122,
219,
294
] | Detected sandwich at location [0, 122, 219, 294] | Detected Sandwich at location [4, 116, 633, 392]
Detected Sandwich at location [2, 119, 402, 364]
Detected Sandwich at location [52, 180, 402, 363]
Detected sandwich at location [0, 122, 219, 294] | [
0,
122,
219,
294
] |
COCO_train2014_000000533220.jpg | refcoco_images/COCO_train2014_000000533220.jpg | right table | [259.03,347.33,594.52,427.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Table'] | Detected Table at location [258, 344, 593, 426]
Detected Table at location [2, 331, 239, 426]
Detected right table at location [258, 346, 594, 426]
| [
258,
346,
594,
426
] | Detected table at location [258, 346, 594, 426] | Detected Table at location [258, 344, 593, 426]
Detected Table at location [2, 331, 239, 426]
Detected table at location [258, 346, 594, 426] | [
258,
346,
594,
426
] |
COCO_train2014_000000537774.jpg | refcoco_images/COCO_train2014_000000537774.jpg | middle lady | [260.73,122.25,423.09,412.58] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Lady'] | Detected Lady at location [258, 120, 424, 424]
Detected middle lady at location [260, 124, 421, 424]
| [
260,
124,
421,
424
] | Detected middle lady at location [260, 124, 421, 424] | Detected Lady at location [258, 120, 424, 424]
Detected middle lady at location [260, 124, 421, 424] | [
260,
124,
421,
424
] |
COCO_train2014_000000537108.jpg | refcoco_images/COCO_train2014_000000537108.jpg | last slice | [248.93,213.17,515.73,380.95] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Slice'] | Detected Slice at location [3, 137, 581, 607]
Detected last slice at location [2, 200, 215, 394]
| [
2,
200,
215,
394
] | Detected last slice at location [2, 200, 215, 394] | Detected Slice at location [3, 137, 581, 607]
Detected last slice at location [2, 200, 215, 394] | [
2,
200,
215,
394
] |
COCO_train2014_000000573166.jpg | refcoco_images/COCO_train2014_000000573166.jpg | left biike | [,83.97,114.9,,326.45,270.84] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Bike'] | Detected Bike at location [84, 109, 478, 272]
Detected Bike at location [85, 112, 327, 271]
Detected left biike at location [85, 113, 329, 273]
| [
85,
113,
329,
273
] | Detected bike at location [85, 113, 329, 273] | Detected Bike at location [84, 109, 478, 272]
Detected Bike at location [85, 112, 327, 271]
Detected bike at location [85, 113, 329, 273] | [
85,
113,
329,
273
] |
COCO_train2014_000000518451.jpg | refcoco_images/COCO_train2014_000000518451.jpg | Woman with both arms out | [126.78,,87.58,345.84,363.73] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Woman'] | Detected Woman at location [246, 36, 424, 363]
Detected Woman with both arms out at location [126, 87, 345, 363]
| [
126,
87,
345,
363
] | Detected Woman at location [126, 87, 345, 363] | Detected Woman at location [246, 36, 424, 363]
Detected Woman at location [126, 87, 345, 363] | [
126,
87,
345,
363
] |
COCO_train2014_000000533944.jpg | refcoco_images/COCO_train2014_000000533944.jpg | BASE PLAYER # 2 | [160.25,119.47,358.11,386.95] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['BASE PLAYER'] | Detected BASE PLAYER at location [156, 113, 356, 389]
Detected BASE PLAYER # 2 at location [157, 119, 354, 387]
| [
157,
119,
354,
387
] | Detected BASE PLAYER at location [157, 119, 354, 387] | Detected BASE PLAYER at location [156, 113, 356, 389]
Detected BASE PLAYER at location [157, 119, 354, 387] | [
157,
119,
354,
387
] |
COCO_train2014_000000580843.jpg | refcoco_images/COCO_train2014_000000580843.jpg | lol: right seat | [402.84,,,0.81,640.,,,426.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Seat'] | Detected Seat at location [35, 0, 639, 422]
Detected Seat at location [36, 0, 378, 423]
Detected lol: right seat at location [403, 0, 639, 425]
| [
403,
0,
639,
425
] | Detected lol at location [403, 0, 639, 425] | Detected Seat at location [35, 0, 639, 422]
Detected Seat at location [36, 0, 378, 423]
Detected lol at location [403, 0, 639, 425] | [
403,
0,
639,
425
] |
COCO_train2014_000000562517.jpg | refcoco_images/COCO_train2014_000000562517.jpg | woman with umbrella | [,27.,,,389.09,231.01,640.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Woman', 'Umbrella'] | Detected Woman at location [29, 248, 297, 639]
Detected umbrella at location [49, 250, 297, 432]
Detected woman with umbrella at location [29, 387, 230, 639]
| [
29,
387,
230,
639
] | Detected woman at location [29, 387, 230, 639] | Detected Woman at location [29, 248, 297, 639]
Detected umbrella at location [49, 250, 297, 432]
Detected woman at location [29, 387, 230, 639] | [
29,
387,
230,
639
] |
COCO_train2014_000000571405.jpg | refcoco_images/COCO_train2014_000000571405.jpg | GUY WITH PINK TOOTHBRUSH | [136.72,,92.09,326.83,355.82] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Guy', 'Pink Toothbrush'] | Detected Guy at location [138, 91, 327, 359]
Detected Toothbrush at location [211, 162, 264, 200]
Detected GUY WITH PINK TOOTHBRUSH at location [473, 107, 591, 359]
| [
473,
107,
591,
359
] | Detected GUY WITH PINK TOOTHBRUSH at location [473, 107, 591, 359] | Detected Guy at location [138, 91, 327, 359]
Detected Toothbrush at location [211, 162, 264, 200]
Detected GUY WITH PINK TOOTHBRUSH at location [473, 107, 591, 359] | [
473,
107,
591,
359
] |
COCO_train2014_000000572245.jpg | refcoco_images/COCO_train2014_000000572245.jpg | left side blk thing | [,,0.,,,,,0.,,,114.34,403.99] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Blk thing'] | Detected Blk thing at location [85, 68, 496, 361]
Detected left side blk thing at location [0, 0, 115, 403]
| [
0,
0,
115,
403
] | Detected blk thing at location [0, 0, 115, 403] | Detected Blk thing at location [85, 68, 496, 361]
Detected blk thing at location [0, 0, 115, 403] | [
0,
0,
115,
403
] |
COCO_train2014_000000550355.jpg | refcoco_images/COCO_train2014_000000550355.jpg | woman on right | [434.43,140.76,639.91,355.96] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Woman'] | Detected Woman at location [435, 138, 639, 359]
Detected woman on right at location [436, 140, 639, 359]
| [
436,
140,
639,
359
] | Detected woman at location [436, 140, 639, 359] | Detected Woman at location [435, 138, 639, 359]
Detected woman at location [436, 140, 639, 359] | [
436,
140,
639,
359
] |
COCO_train2014_000000547533.jpg | refcoco_images/COCO_train2014_000000547533.jpg | truck in back | [380.4,,138.57,480.,,,358.84] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Truck'] | Detected Truck at location [1, 259, 252, 639]
Detected truck in back at location [0, 260, 251, 639]
| [
0,
260,
251,
639
] | Detected truck at location [0, 260, 251, 639] | Detected Truck at location [1, 259, 252, 639]
Detected truck at location [0, 260, 251, 639] | [
0,
260,
251,
639
] |
COCO_train2014_000000544449.jpg | refcoco_images/COCO_train2014_000000544449.jpg | top right corner apple | [462.14,,,0.,,,640.,,,108.55] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Apple'] | Detected Apple at location [3, 2, 638, 463]
Detected Apple at location [19, 225, 244, 454]
Detected Apple at location [42, 39, 249, 240]
Detected top right corner apple at location [464, 0, 638, 105]
| [
464,
0,
638,
105
] | Detected apple at location [464, 0, 638, 105] | Detected Apple at location [3, 2, 638, 463]
Detected Apple at location [19, 225, 244, 454]
Detected Apple at location [42, 39, 249, 240]
Detected apple at location [464, 0, 638, 105] | [
464,
0,
638,
105
] |
COCO_train2014_000000520094.jpg | refcoco_images/COCO_train2014_000000520094.jpg | guy in back right | [330.04,133.32,450.57,429.17] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Guy'] | Detected Guy at location [84, 16, 292, 639]
Detected guy in back right at location [329, 133, 450, 429]
| [
329,
133,
450,
429
] | Detected guy at location [329, 133, 450, 429] | Detected Guy at location [84, 16, 292, 639]
Detected guy at location [329, 133, 450, 429] | [
329,
133,
450,
429
] |
COCO_train2014_000000542988.jpg | refcoco_images/COCO_train2014_000000542988.jpg | back zebra | [178.48,127.65,530.71,394.38] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Zebra'] | Detected Zebra at location [18, 102, 535, 396]
Detected Zebra at location [16, 105, 335, 396]
Detected Zebra at location [181, 126, 532, 390]
Detected back zebra at location [178, 128, 530, 392]
| [
178,
128,
530,
392
] | Detected zebra at location [178, 128, 530, 392] | Detected Zebra at location [18, 102, 535, 396]
Detected Zebra at location [16, 105, 335, 396]
Detected Zebra at location [181, 126, 532, 390]
Detected zebra at location [178, 128, 530, 392] | [
178,
128,
530,
392
] |
COCO_train2014_000000569946.jpg | refcoco_images/COCO_train2014_000000569946.jpg | right woman white dress | [414.56,161.22,557.01,441.24] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Woman', 'White dress'] | Detected Woman at location [415, 159, 558, 443]
Detected Woman at location [210, 138, 331, 447]
Detected dress at location [415, 207, 557, 441]
Detected dress at location [211, 191, 331, 443]
Detected right woman white dress at location [415, 161, 556, 440]
| [
415,
161,
556,
440
] | Detected woman in white dress at location [415, 161, 556, 440] | Detected Woman at location [415, 159, 558, 443]
Detected Woman at location [210, 138, 331, 447]
Detected dress at location [415, 207, 557, 441]
Detected dress at location [211, 191, 331, 443]
Detected woman in white dress at location [415, 161, 556, 440] | [
415,
161,
556,
440
] |
COCO_train2014_000000541553.jpg | refcoco_images/COCO_train2014_000000541553.jpg | blond dog | [,,0.,,,249.27,263.66,474.3,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Blond', 'Dog'] | Detected Blond at location [265, 37, 465, 291]
Detected Dog at location [265, 37, 464, 291]
Detected blond dog at location [0, 250, 263, 479]
| [
0,
250,
263,
479
] | Detected blond dog at location [0, 250, 263, 479] | Detected Blond at location [265, 37, 465, 291]
Detected Dog at location [265, 37, 464, 291]
Detected blond dog at location [0, 250, 263, 479] | [
0,
250,
263,
479
] |
COCO_train2014_000000529136.jpg | refcoco_images/COCO_train2014_000000529136.jpg | first bike in front | [,,0.,,,216.29,374.82,480.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Bike'] | Detected Bike at location [4, 162, 516, 479]
Detected Bike at location [3, 212, 373, 479]
Detected first bike in front at location [0, 212, 373, 479]
| [
0,
212,
373,
479
] | Detected bike at location [0, 212, 373, 479] | Detected Bike at location [4, 162, 516, 479]
Detected Bike at location [3, 212, 373, 479]
Detected bike at location [0, 212, 373, 479] | [
0,
212,
373,
479
] |
COCO_train2014_000000562517.jpg | refcoco_images/COCO_train2014_000000562517.jpg | Woman in red coat | [397.55,248.93,515.22,595.79] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Woman', 'Red coat', 'Coat'] | Detected Woman at location [394, 248, 512, 599]
Detected coat at location [395, 288, 508, 448]
Detected Woman in red coat at location [395, 249, 514, 596]
| [
395,
249,
514,
596
] | Detected Woman in red coat at location [395, 249, 514, 596] | Detected Woman at location [394, 248, 512, 599]
Detected coat at location [395, 288, 508, 448]
Detected Woman in red coat at location [395, 249, 514, 596] | [
395,
249,
514,
596
] |
COCO_train2014_000000564063.jpg | refcoco_images/COCO_train2014_000000564063.jpg | truck closest to us | [,15.38,138.15,598.63,510.63] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Truck'] | Detected Truck at location [15, 140, 619, 510]
Detected truck closest to us at location [17, 139, 597, 510]
| [
17,
139,
597,
510
] | Detected truck at location [17, 139, 597, 510] | Detected Truck at location [15, 140, 619, 510]
Detected truck at location [17, 139, 597, 510] | [
17,
139,
597,
510
] |
COCO_train2014_000000545549.jpg | refcoco_images/COCO_train2014_000000545549.jpg | guy right | [540.47,181.83,629.04,440.67] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Guy'] | Detected Guy at location [181, 194, 349, 476]
Detected guy right at location [540, 182, 628, 439]
| [
540,
182,
628,
439
] | Detected guy at location [540, 182, 628, 439] | Detected Guy at location [181, 194, 349, 476]
Detected guy at location [540, 182, 628, 439] | [
540,
182,
628,
439
] |
COCO_train2014_000000565652.jpg | refcoco_images/COCO_train2014_000000565652.jpg | zebra on left | [,39.43,102.42,285.63,357.28] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Zebra'] | Detected Zebra at location [39, 93, 604, 360]
Detected Zebra at location [272, 118, 602, 357]
Detected Zebra at location [130, 96, 389, 354]
Detected Zebra at location [41, 101, 284, 355]
Detected zebra on left at location [40, 104, 283, 355]
| [
40,
104,
283,
355
] | Detected zebra at location [40, 104, 283, 355] | Detected Zebra at location [39, 93, 604, 360]
Detected Zebra at location [272, 118, 602, 357]
Detected Zebra at location [130, 96, 389, 354]
Detected Zebra at location [41, 101, 284, 355]
Detected zebra at location [40, 104, 283, 355] | [
40,
104,
283,
355
] |
COCO_train2014_000000541450.jpg | refcoco_images/COCO_train2014_000000541450.jpg | broccoli bottom left of bowl | [,24.91,290.94,270.57,455.09] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Broccoli', 'Bowl'] | Detected Broccoli at location [152, 125, 352, 309]
Detected broccoli bottom left of bowl at location [25, 292, 270, 455]
| [
25,
292,
270,
455
] | Detected broccoli at location [25, 292, 270, 455] | Detected Broccoli at location [152, 125, 352, 309]
Detected broccoli at location [25, 292, 270, 455] | [
25,
292,
270,
455
] |
COCO_train2014_000000562826.jpg | refcoco_images/COCO_train2014_000000562826.jpg | RIGHT BEHIND HEAD OF MAN DRNKING | [1.3915e+02,2.7000e-01,3.9155e+02,1.7825e+02] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Head', 'Man', 'Drink'] | Detected Head at location [247, 12, 409, 210]
Detected Man at location [112, 12, 565, 471]
| [
112,
12,
565,
471
] | Detected Man at location [112, 12, 565, 471] | Detected Head at location [247, 12, 409, 210]
Detected Man at location [112, 12, 565, 471] | [
112,
12,
565,
471
] |
COCO_train2014_000000563525.jpg | refcoco_images/COCO_train2014_000000563525.jpg | woman sitting | [132.69,,43.09,335.21,324.43] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Woman'] | Detected Woman at location [129, 40, 338, 325]
Detected woman sitting at location [133, 44, 334, 324]
| [
133,
44,
334,
324
] | Detected woman at location [133, 44, 334, 324] | Detected Woman at location [129, 40, 338, 325]
Detected woman at location [133, 44, 334, 324] | [
133,
44,
334,
324
] |
COCO_train2014_000000560309.jpg | refcoco_images/COCO_train2014_000000560309.jpg | the chair on the right | [193.37,240.29,331.02,395.39] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Chair'] | Detected Chair at location [51, 239, 199, 399]
Detected Chair at location [195, 239, 332, 396]
Detected the chair on the right at location [193, 241, 332, 396]
| [
193,
241,
332,
396
] | Detected chair at location [193, 241, 332, 396] | Detected Chair at location [51, 239, 199, 399]
Detected Chair at location [195, 239, 332, 396]
Detected chair at location [193, 241, 332, 396] | [
193,
241,
332,
396
] |
COCO_train2014_000000535418.jpg | refcoco_images/COCO_train2014_000000535418.jpg | purple cap | [113.01,166.28,314.33,481.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Purple cap'] | Detected Purple cap at location [114, 166, 312, 338]
Detected purple cap at location [115, 167, 315, 342]
| [
115,
167,
315,
342
] | Detected purple cap at location [115, 167, 315, 342] | Detected Purple cap at location [114, 166, 312, 338]
Detected purple cap at location [115, 167, 315, 342] | [
115,
167,
315,
342
] |
COCO_train2014_000000565311.jpg | refcoco_images/COCO_train2014_000000565311.jpg | center piece on left is blurry | [,52.56,110.69,214.35,224.04] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Center piece'] | Detected Center piece at location [292, 142, 638, 392]
Detected center piece on left is blurry at location [52, 111, 214, 224]
| [
52,
111,
214,
224
] | Detected center piece at location [52, 111, 214, 224] | Detected Center piece at location [292, 142, 638, 392]
Detected center piece at location [52, 111, 214, 224] | [
52,
111,
214,
224
] |
COCO_train2014_000000534543.jpg | refcoco_images/COCO_train2014_000000534543.jpg | middle child dark hair | [232.07,,50.45,409.37,557.84] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Child', 'Hair'] | Detected Child at location [300, 71, 569, 638]
Detected middle child dark hair at location [232, 50, 408, 556]
| [
232,
50,
408,
556
] | Detected middle child at location [232, 50, 408, 556] | Detected Child at location [300, 71, 569, 638]
Detected middle child at location [232, 50, 408, 556] | [
232,
50,
408,
556
] |
COCO_train2014_000000526204.jpg | refcoco_images/COCO_train2014_000000526204.jpg | man standing white shirt | [291.93,,31.87,385.13,311.47] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Man', 'White shirt'] | Detected Man at location [291, 32, 386, 312]
Detected shirt at location [298, 86, 386, 240]
Detected man standing white shirt at location [291, 32, 383, 309]
| [
291,
32,
383,
309
] | Detected man at location [291, 32, 383, 309] | Detected Man at location [291, 32, 386, 312]
Detected shirt at location [298, 86, 386, 240]
Detected man at location [291, 32, 383, 309] | [
291,
32,
383,
309
] |
COCO_train2014_000000573854.jpg | refcoco_images/COCO_train2014_000000573854.jpg | smaller, greenish plane | [195.74,228.03,533.7,,328.01] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Plane'] | Detected Plane at location [45, 106, 600, 275]
Detected smaller, greenish plane at location [196, 229, 533, 328]
| [
196,
229,
533,
328
] | Detected plane at location [196, 229, 533, 328] | Detected Plane at location [45, 106, 600, 275]
Detected plane at location [196, 229, 533, 328] | [
196,
229,
533,
328
] |
COCO_train2014_000000545405.jpg | refcoco_images/COCO_train2014_000000545405.jpg | hand holding the hot dog | [,58.15,183.76,271.35,367.47] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Hand', 'Hot Dog'] | Detected Hand at location [249, 51, 463, 466]
Detected Hot Dog at location [249, 51, 462, 466]
Detected hand holding the hot dog at location [68, 228, 358, 479]
| [
68,
228,
358,
479
] | Detected hot dog at location [68, 228, 358, 479] | Detected Hand at location [249, 51, 463, 466]
Detected Hot Dog at location [249, 51, 462, 466]
Detected hot dog at location [68, 228, 358, 479] | [
68,
228,
358,
479
] |
COCO_train2014_000000531366.jpg | refcoco_images/COCO_train2014_000000531366.jpg | calf in back | [469.1,,170.13,617.24,312.09] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Calf'] | Detected Calf at location [325, 166, 620, 313]
Detected Calf at location [328, 175, 492, 313]
Detected calf in back at location [459, 169, 619, 310]
| [
459,
169,
619,
310
] | Detected calf at location [459, 169, 619, 310] | Detected Calf at location [325, 166, 620, 313]
Detected Calf at location [328, 175, 492, 313]
Detected calf at location [459, 169, 619, 310] | [
459,
169,
619,
310
] |
COCO_train2014_000000517460.jpg | refcoco_images/COCO_train2014_000000517460.jpg | player on right | [133.15,152.59,346.85,395.27] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Player'] | Detected Player at location [35, 160, 272, 394]
Detected player on right at location [133, 153, 346, 394]
| [
133,
153,
346,
394
] | Detected player at location [133, 153, 346, 394] | Detected Player at location [35, 160, 272, 394]
Detected player at location [133, 153, 346, 394] | [
133,
153,
346,
394
] |
COCO_train2014_000000563929.jpg | refcoco_images/COCO_train2014_000000563929.jpg | middle zebra | [210.82,109.21,370.96,427.46] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Zebra'] | Detected Zebra at location [74, 85, 500, 421]
Detected middle zebra at location [211, 109, 369, 425]
| [
211,
109,
369,
425
] | Detected zebra at location [211, 109, 369, 425] | Detected Zebra at location [74, 85, 500, 421]
Detected zebra at location [211, 109, 369, 425] | [
211,
109,
369,
425
] |
COCO_train2014_000000578718.jpg | refcoco_images/COCO_train2014_000000578718.jpg | pink shirt | [132.79,301.64,437.6,,640.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Pink shirt', 'Shirt'] | Detected Pink shirt at location [183, 431, 441, 639]
Detected pink shirt at location [183, 432, 443, 639]
| [
183,
432,
443,
639
] | Detected pink shirt at location [183, 432, 443, 639] | Detected Pink shirt at location [183, 431, 441, 639]
Detected pink shirt at location [183, 432, 443, 639] | [
183,
432,
443,
639
] |
COCO_train2014_000000533525.jpg | refcoco_images/COCO_train2014_000000533525.jpg | middle zebra | [139.26,140.98,376.63,360.25] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Zebra'] | Detected Zebra at location [84, 61, 598, 362]
Detected Zebra at location [139, 142, 378, 361]
Detected Zebra at location [430, 87, 597, 306]
Detected Zebra at location [87, 63, 234, 261]
Detected middle zebra at location [139, 143, 377, 361]
| [
139,
143,
377,
361
] | Detected middle zebra at location [139, 143, 377, 361] | Detected Zebra at location [84, 61, 598, 362]
Detected Zebra at location [139, 142, 378, 361]
Detected Zebra at location [430, 87, 597, 306]
Detected Zebra at location [87, 63, 234, 261]
Detected middle zebra at location [139, 143, 377, 361] | [
139,
143,
377,
361
] |
COCO_train2014_000000546218.jpg | refcoco_images/COCO_train2014_000000546218.jpg | left glass | [,46.33,,78.05,174.22,296.58] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Glass'] | Detected Glass at location [43, 0, 606, 307]
Detected Glass at location [44, 42, 604, 306]
Detected left glass at location [47, 78, 173, 295]
| [
47,
78,
173,
295
] | Detected glass at location [47, 78, 173, 295] | Detected Glass at location [43, 0, 606, 307]
Detected Glass at location [44, 42, 604, 306]
Detected glass at location [47, 78, 173, 295] | [
47,
78,
173,
295
] |
COCO_train2014_000000537326.jpg | refcoco_images/COCO_train2014_000000537326.jpg | left man | [138.15,,38.64,327.82,465.97] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Man'] | Detected Man at location [153, 35, 327, 460]
Detected left man at location [141, 39, 327, 463]
| [
141,
39,
327,
463
] | Detected man at location [141, 39, 327, 463] | Detected Man at location [153, 35, 327, 460]
Detected man at location [141, 39, 327, 463] | [
141,
39,
327,
463
] |
COCO_train2014_000000535820.jpg | refcoco_images/COCO_train2014_000000535820.jpg | lady in dress | [185.74,,61.23,290.09,376.21] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Lady', 'Dress'] | Detected Lady at location [187, 60, 289, 377]
Detected Dress at location [201, 112, 277, 302]
Detected lady in dress at location [187, 62, 290, 376]
| [
187,
62,
290,
376
] | Detected lady at location [187, 62, 290, 376] | Detected Lady at location [187, 60, 289, 377]
Detected Dress at location [201, 112, 277, 302]
Detected lady at location [187, 62, 290, 376] | [
187,
62,
290,
376
] |
COCO_train2014_000000522461.jpg | refcoco_images/COCO_train2014_000000522461.jpg | THE GIRL THAT IS ON THE FLOOR IN ALL WHITE | [119.17,231.87,408.03,470.21] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Girl', 'White'] | Detected Girl at location [115, 234, 407, 478]
Detected THE GIRL THAT IS ON THE FLOOR IN ALL WHITE at location [119, 234, 406, 475]
| [
119,
234,
406,
475
] | Detected girl in all white at location [119, 234, 406, 475] | Detected Girl at location [115, 234, 407, 478]
Detected girl in all white at location [119, 234, 406, 475] | [
119,
234,
406,
475
] |
COCO_train2014_000000563592.jpg | refcoco_images/COCO_train2014_000000563592.jpg | very left cow | [,,0.96,136.26,146.81,402.05] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Cow'] | Detected Cow at location [389, 130, 639, 422]
Detected Cow at location [2, 93, 293, 329]
Detected Cow at location [2, 139, 144, 400]
Detected very left cow at location [0, 137, 144, 401]
| [
0,
137,
144,
401
] | Detected cow at location [0, 137, 144, 401] | Detected Cow at location [389, 130, 639, 422]
Detected Cow at location [2, 93, 293, 329]
Detected Cow at location [2, 139, 144, 400]
Detected cow at location [0, 137, 144, 401] | [
0,
137,
144,
401
] |
COCO_train2014_000000551575.jpg | refcoco_images/COCO_train2014_000000551575.jpg | yellow and blue dudes board | [334.72,263.93,583.48,387.3,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Dude', 'Board'] | Detected Dude at location [0, 74, 584, 419]
Detected Dude at location [0, 126, 582, 416]
Detected Board at location [0, 128, 182, 419]
Detected yellow and blue dudes board at location [0, 124, 186, 421]
| [
0,
124,
186,
421
] | Detected yellow dudes at location [0, 124, 186, 421] | Detected Dude at location [0, 74, 584, 419]
Detected Dude at location [0, 126, 582, 416]
Detected Board at location [0, 128, 182, 419]
Detected yellow dudes at location [0, 124, 186, 421] | [
0,
124,
186,
421
] |
COCO_train2014_000000523937.jpg | refcoco_images/COCO_train2014_000000523937.jpg | back of guy with bright colors | [216.7,,,97.25,382.48,474.31] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Guy', 'Colors'] | Detected Guy at location [219, 78, 372, 479]
Detected back of guy with bright colors at location [220, 79, 370, 479]
| [
220,
79,
370,
479
] | Detected guy at location [220, 79, 370, 479] | Detected Guy at location [219, 78, 372, 479]
Detected guy at location [220, 79, 370, 479] | [
220,
79,
370,
479
] |
COCO_train2014_000000534633.jpg | refcoco_images/COCO_train2014_000000534633.jpg | skiier on left | [129.76,141.31,342.39,436.01] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Skiier'] | Detected Skiier at location [128, 139, 332, 438]
Detected skiier on left at location [130, 142, 328, 436]
| [
130,
142,
328,
436
] | Detected skiier at location [130, 142, 328, 436] | Detected Skiier at location [128, 139, 332, 438]
Detected skiier at location [130, 142, 328, 436] | [
130,
142,
328,
436
] |
COCO_train2014_000000535874.jpg | refcoco_images/COCO_train2014_000000535874.jpg | left chair | [108.39,221.93,286.03,396.28] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Chair'] | Detected Chair at location [107, 221, 288, 395]
Detected left chair at location [108, 222, 285, 395]
| [
108,
222,
285,
395
] | Detected chair at location [108, 222, 285, 395] | Detected Chair at location [107, 221, 288, 395]
Detected chair at location [108, 222, 285, 395] | [
108,
222,
285,
395
] |
COCO_train2014_000000538149.jpg | refcoco_images/COCO_train2014_000000538149.jpg | second bear from right | [402.75,209.92,532.25,421.4,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Bear'] | Detected Bear at location [14, 15, 639, 460]
Detected second bear from right at location [403, 210, 532, 421]
| [
403,
210,
532,
421
] | Detected bear at location [403, 210, 532, 421] | Detected Bear at location [14, 15, 639, 460]
Detected bear at location [403, 210, 532, 421] | [
403,
210,
532,
421
] |
COCO_train2014_000000579533.jpg | refcoco_images/COCO_train2014_000000579533.jpg | white soilder | [,21.84,,23.51,236.15,457.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Soldier'] | Detected Soldier at location [190, 21, 481, 456]
Detected Soldier at location [21, 23, 236, 456]
Detected white soilder at location [22, 24, 235, 456]
| [
22,
24,
235,
456
] | Detected white soldier at location [22, 24, 235, 456] | Detected Soldier at location [190, 21, 481, 456]
Detected Soldier at location [21, 23, 236, 456]
Detected white soldier at location [22, 24, 235, 456] | [
22,
24,
235,
456
] |
COCO_train2014_000000543413.jpg | refcoco_images/COCO_train2014_000000543413.jpg | baby elephant | [379.13,243.29,497.85,409.88] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Baby', 'Elephant'] | Detected Baby at location [29, 134, 510, 420]
Detected elephant at location [30, 137, 395, 421]
Detected baby elephant at location [380, 244, 498, 409]
| [
380,
244,
498,
409
] | Detected baby elephant at location [380, 244, 498, 409] | Detected Baby at location [29, 134, 510, 420]
Detected elephant at location [30, 137, 395, 421]
Detected baby elephant at location [380, 244, 498, 409] | [
380,
244,
498,
409
] |
COCO_train2014_000000579051.jpg | refcoco_images/COCO_train2014_000000579051.jpg | ki | [288.75,100.98,435.61,622.8,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | [] | Detected ki at location [62, 404, 130, 516]
| [
62,
404,
130,
516
] | Detected ki at location [62, 404, 130, 516] | Detected ki at location [62, 404, 130, 516] | [
62,
404,
130,
516
] |
COCO_train2014_000000561655.jpg | refcoco_images/COCO_train2014_000000561655.jpg | hair on left | [,,0.,,,,30.29,311.59,422.67] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Hair'] | Detected Hair at location [3, 27, 320, 423]
Detected Hair at location [3, 133, 318, 423]
Detected hair on left at location [0, 31, 311, 422]
| [
0,
31,
311,
422
] | Detected hair at location [0, 31, 311, 422] | Detected Hair at location [3, 27, 320, 423]
Detected Hair at location [3, 133, 318, 423]
Detected hair at location [0, 31, 311, 422] | [
0,
31,
311,
422
] |
COCO_train2014_000000573611.jpg | refcoco_images/COCO_train2014_000000573611.jpg | far left guy | [,,0.,,,,81.85,142.79,409.06] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Guy'] | Detected Guy at location [4, 101, 500, 639]
Detected far left guy at location [0, 82, 142, 409]
| [
0,
82,
142,
409
] | Detected guy at location [0, 82, 142, 409] | Detected Guy at location [4, 101, 500, 639]
Detected guy at location [0, 82, 142, 409] | [
0,
82,
142,
409
] |
COCO_train2014_000000578567.jpg | refcoco_images/COCO_train2014_000000578567.jpg | Pitcher. | [,96.14,172.13,284.13,625.58] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Pitcher'] | Detected Pitcher at location [94, 170, 290, 627]
Detected Pitcher. at location [96, 172, 284, 622]
| [
96,
172,
284,
622
] | Detected Pitcher at location [96, 172, 284, 622] | Detected Pitcher at location [94, 170, 290, 627]
Detected Pitcher at location [96, 172, 284, 622] | [
96,
172,
284,
622
] |
COCO_train2014_000000545944.jpg | refcoco_images/COCO_train2014_000000545944.jpg | person to the left | [,,1.07,,57.41,171.81,474.73] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Person'] | Detected Person at location [367, 2, 639, 479]
Detected person to the left at location [0, 57, 172, 479]
| [
0,
57,
172,
479
] | Detected person at location [0, 57, 172, 479] | Detected Person at location [367, 2, 639, 479]
Detected person at location [0, 57, 172, 479] | [
0,
57,
172,
479
] |
COCO_train2014_000000544113.jpg | refcoco_images/COCO_train2014_000000544113.jpg | vase holding flowers | [169.37,368.09,228.16,588.34] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Vase', 'Flowers'] | Detected Vase at location [167, 367, 225, 588]
Detected Flowers at location [3, 81, 361, 504]
Detected vase holding flowers at location [169, 368, 224, 588]
| [
169,
368,
224,
588
] | Detected vase at location [169, 368, 224, 588] | Detected Vase at location [167, 367, 225, 588]
Detected Flowers at location [3, 81, 361, 504]
Detected vase at location [169, 368, 224, 588] | [
169,
368,
224,
588
] |
COCO_train2014_000000581057.jpg | refcoco_images/COCO_train2014_000000581057.jpg | main bike | [,77.57,,62.69,515.46,414.53] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Bike'] | Detected Bike at location [75, 61, 517, 414]
Detected main bike at location [77, 62, 514, 415]
| [
77,
62,
514,
415
] | Detected main bike at location [77, 62, 514, 415] | Detected Bike at location [75, 61, 517, 414]
Detected main bike at location [77, 62, 514, 415] | [
77,
62,
514,
415
] |
COCO_train2014_000000544018.jpg | refcoco_images/COCO_train2014_000000544018.jpg | black coffee cup | [,39.25,,65.15,275.52,312.8,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Coffee cup'] | Detected Coffee cup at location [33, 55, 344, 364]
Detected Coffee cup at location [38, 60, 277, 308]
Detected black coffee cup at location [40, 65, 273, 308]
| [
40,
65,
273,
308
] | Detected black coffee cup at location [40, 65, 273, 308] | Detected Coffee cup at location [33, 55, 344, 364]
Detected Coffee cup at location [38, 60, 277, 308]
Detected black coffee cup at location [40, 65, 273, 308] | [
40,
65,
273,
308
] |
COCO_train2014_000000564950.jpg | refcoco_images/COCO_train2014_000000564950.jpg | blue shirt | [,,0.,,,,20.2,,142.81,431.09] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Blue shirt', 'Shirt'] | Detected Blue shirt at location [4, 0, 425, 639]
Detected blue shirt at location [0, 19, 142, 427]
| [
0,
19,
142,
427
] | Detected blue shirt at location [0, 19, 142, 427] | Detected Blue shirt at location [4, 0, 425, 639]
Detected blue shirt at location [0, 19, 142, 427] | [
0,
19,
142,
427
] |
COCO_train2014_000000545253.jpg | refcoco_images/COCO_train2014_000000545253.jpg | right bench | [298.92,100.56,516.44,234.84] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Bench'] | Detected Bench at location [73, 83, 517, 416]
Detected Bench at location [75, 86, 430, 416]
Detected right bench at location [300, 100, 516, 234]
| [
300,
100,
516,
234
] | Detected bench at location [300, 100, 516, 234] | Detected Bench at location [73, 83, 517, 416]
Detected Bench at location [75, 86, 430, 416]
Detected bench at location [300, 100, 516, 234] | [
300,
100,
516,
234
] |
COCO_train2014_000000556162.jpg | refcoco_images/COCO_train2014_000000556162.jpg | boy holding teddy bear on right | [279.37,,72.26,636.87,480.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Boy', 'Teddy bear'] | Detected Boy at location [19, 70, 639, 479]
Detected Boy at location [276, 71, 639, 479]
Detected boy holding teddy bear on right at location [278, 73, 639, 479]
| [
278,
73,
639,
479
] | Detected boy at location [278, 73, 639, 479] | Detected Boy at location [19, 70, 639, 479]
Detected Boy at location [276, 71, 639, 479]
Detected boy at location [278, 73, 639, 479] | [
278,
73,
639,
479
] |
COCO_train2014_000000525101.jpg | refcoco_images/COCO_train2014_000000525101.jpg | train almost out of pic | [,,0.,,,334.65,138.99,639.95] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Train'] | Detected Train at location [166, 272, 278, 421]
Detected train almost out of pic at location [166, 272, 278, 422]
| [
166,
272,
278,
422
] | Detected train at location [166, 272, 278, 422] | Detected Train at location [166, 272, 278, 421]
Detected train at location [166, 272, 278, 422] | [
166,
272,
278,
422
] |
COCO_train2014_000000570836.jpg | refcoco_images/COCO_train2014_000000570836.jpg | The tall vase. | [138.07,321.08,290.52,634.61] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Vase'] | Detected Vase at location [140, 321, 289, 636]
Detected The tall vase. at location [141, 323, 288, 636]
| [
141,
323,
288,
636
] | Detected vase at location [141, 323, 288, 636] | Detected Vase at location [140, 321, 289, 636]
Detected vase at location [141, 323, 288, 636] | [
141,
323,
288,
636
] |
COCO_train2014_000000526514.jpg | refcoco_images/COCO_train2014_000000526514.jpg | car on the far left with little yellow light | [,,0.69,234.06,141.54,428.36] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Car', 'Light'] | Detected Car at location [291, 254, 412, 308]
Detected car on the far left with little yellow light at location [0, 237, 140, 428]
| [
0,
237,
140,
428
] | Detected car at location [0, 237, 140, 428] | Detected Car at location [291, 254, 412, 308]
Detected car at location [0, 237, 140, 428] | [
0,
237,
140,
428
] |
COCO_train2014_000000549003.jpg | refcoco_images/COCO_train2014_000000549003.jpg | bottom right | [526.5,,202.48,640.,,,427.,,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | [] | Detected bottom right at location [523, 200, 639, 426]
| [
523,
200,
639,
426
] | Detected at location [523, 200, 639, 426] | Detected at location [523, 200, 639, 426] | [
523,
200,
639,
426
] |
COCO_train2014_000000564715.jpg | refcoco_images/COCO_train2014_000000564715.jpg | dude with blue glasses | [365.,,,286.15,549.64,469.72] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Dude', 'Glasses'] | Detected Dude at location [365, 286, 550, 474]
Detected glasses at location [435, 323, 498, 355]
Detected dude with blue glasses at location [366, 289, 547, 474]
| [
366,
289,
547,
474
] | Detected dude with blue glasses at location [366, 289, 547, 474] | Detected Dude at location [365, 286, 550, 474]
Detected glasses at location [435, 323, 498, 355]
Detected dude with blue glasses at location [366, 289, 547, 474] | [
366,
289,
547,
474
] |
COCO_train2014_000000529624.jpg | refcoco_images/COCO_train2014_000000529624.jpg | woman on left side in front of net | [136.63,198.47,294.83,632.81] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Woman', 'Net'] | Detected Woman at location [139, 197, 318, 639]
Detected Woman at location [270, 220, 434, 639]
Detected Net at location [0, 398, 639, 639]
Detected woman on left side in front of net at location [140, 199, 317, 639]
| [
140,
199,
317,
639
] | Detected woman at location [140, 199, 317, 639] | Detected Woman at location [139, 197, 318, 639]
Detected Woman at location [270, 220, 434, 639]
Detected Net at location [0, 398, 639, 639]
Detected woman at location [140, 199, 317, 639] | [
140,
199,
317,
639
] |
COCO_train2014_000000543838.jpg | refcoco_images/COCO_train2014_000000543838.jpg | center blue vase | [214.35,135.04,391.19,390.12] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Blue vase'] | Detected Blue vase at location [213, 131, 394, 394]
Detected Blue vase at location [25, 128, 164, 396]
Detected center blue vase at location [215, 133, 392, 393]
| [
215,
133,
392,
393
] | Detected blue vase at location [215, 133, 392, 393] | Detected Blue vase at location [213, 131, 394, 394]
Detected Blue vase at location [25, 128, 164, 396]
Detected blue vase at location [215, 133, 392, 393] | [
215,
133,
392,
393
] |
COCO_train2014_000000576128.jpg | refcoco_images/COCO_train2014_000000576128.jpg | Right elephant | [426.07,139.22,627.78,317.2,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Elephant'] | Detected Elephant at location [85, 167, 420, 385]
Detected Right elephant at location [427, 139, 627, 317]
| [
427,
139,
627,
317
] | Detected elephant at location [427, 139, 627, 317] | Detected Elephant at location [85, 167, 420, 385]
Detected elephant at location [427, 139, 627, 317] | [
427,
139,
627,
317
] |
COCO_train2014_000000532269.jpg | refcoco_images/COCO_train2014_000000532269.jpg | player on right | [256.03,,,1.69,519.47,402.64] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Player'] | Detected Player at location [255, 0, 520, 403]
Detected player on right at location [256, 0, 518, 402]
| [
256,
0,
518,
402
] | Detected player at location [256, 0, 518, 402] | Detected Player at location [255, 0, 520, 403]
Detected player at location [256, 0, 518, 402] | [
256,
0,
518,
402
] |
COCO_train2014_000000569795.jpg | refcoco_images/COCO_train2014_000000569795.jpg | guys reflection | [,15.24,142.,,,194.7,,369.71] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Guys', 'Reflection'] | Detected Guys at location [18, 33, 639, 479]
Detected Guys at location [20, 180, 248, 479]
Detected reflection at location [20, 181, 247, 479]
Detected guys reflection at location [20, 181, 248, 479]
| [
20,
181,
248,
479
] | Detected guys at location [20, 181, 248, 479] | Detected Guys at location [18, 33, 639, 479]
Detected Guys at location [20, 180, 248, 479]
Detected reflection at location [20, 181, 247, 479]
Detected guys at location [20, 181, 248, 479] | [
20,
181,
248,
479
] |
COCO_train2014_000000563839.jpg | refcoco_images/COCO_train2014_000000563839.jpg | grey shirt | [319.31,,,1.05,510.89,416.67] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Grey shirt', 'Shirt'] | Detected Grey shirt at location [319, 62, 510, 420]
Detected grey shirt at location [319, 0, 509, 419]
| [
319,
0,
509,
419
] | Detected grey shirt at location [319, 0, 509, 419] | Detected Grey shirt at location [319, 62, 510, 420]
Detected grey shirt at location [319, 0, 509, 419] | [
319,
0,
509,
419
] |
COCO_train2014_000000559618.jpg | refcoco_images/COCO_train2014_000000559618.jpg | pc on right | [308.06,117.88,612.,,,472.71] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['PC'] | Detected PC at location [309, 107, 611, 473]
Detected PC at location [0, 108, 289, 466]
Detected pc on right at location [309, 110, 611, 471]
| [
309,
110,
611,
471
] | Detected pc at location [309, 110, 611, 471] | Detected PC at location [309, 107, 611, 473]
Detected PC at location [0, 108, 289, 466]
Detected pc at location [309, 110, 611, 471] | [
309,
110,
611,
471
] |
COCO_train2014_000000530097.jpg | refcoco_images/COCO_train2014_000000530097.jpg | center laptop on bottom | [120.49,,,5.16,458.74,363.2,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Laptop'] | Detected Laptop at location [120, 9, 455, 362]
Detected Laptop at location [150, 60, 380, 298]
Detected center laptop on bottom at location [150, 62, 379, 297]
| [
150,
62,
379,
297
] | Detected laptop at location [150, 62, 379, 297] | Detected Laptop at location [120, 9, 455, 362]
Detected Laptop at location [150, 60, 380, 298]
Detected laptop at location [150, 62, 379, 297] | [
150,
62,
379,
297
] |
COCO_train2014_000000554442.jpg | refcoco_images/COCO_train2014_000000554442.jpg | zebra on the right in front | [270.83,187.3,,575.6,,416.93] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Zebra'] | Detected Zebra at location [56, 177, 578, 424]
Detected Zebra at location [270, 179, 575, 415]
Detected Zebra at location [59, 216, 359, 425]
Detected zebra on the right in front at location [271, 187, 575, 414]
| [
271,
187,
575,
414
] | Detected zebra at location [271, 187, 575, 414] | Detected Zebra at location [56, 177, 578, 424]
Detected Zebra at location [270, 179, 575, 415]
Detected Zebra at location [59, 216, 359, 425]
Detected zebra at location [271, 187, 575, 414] | [
271,
187,
575,
414
] |
COCO_train2014_000000577808.jpg | refcoco_images/COCO_train2014_000000577808.jpg | RIGHT HALF OF SANDWHICH | [275.81,,85.08,587.3,,385.59] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Sandwich'] | Detected Sandwich at location [17, 46, 578, 386]
| [
17,
46,
578,
386
] | Detected Sandwich at location [17, 46, 578, 386] | Detected Sandwich at location [17, 46, 578, 386] | [
17,
46,
578,
386
] |
COCO_train2014_000000548393.jpg | refcoco_images/COCO_train2014_000000548393.jpg | top left donut | [,69.33,,50.56,200.57,185.25] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Donut'] | Detected Donut at location [64, 47, 448, 325]
Detected top left donut at location [69, 50, 200, 184]
| [
69,
50,
200,
184
] | Detected donut at location [69, 50, 200, 184] | Detected Donut at location [64, 47, 448, 325]
Detected donut at location [69, 50, 200, 184] | [
69,
50,
200,
184
] |
COCO_train2014_000000550298.jpg | refcoco_images/COCO_train2014_000000550298.jpg | green apple, top center frame | [242.16,,,0.54,387.03,134.59] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Green apple'] | Detected Green apple at location [389, 0, 575, 179]
Detected Green apple at location [245, 0, 385, 135]
Detected green apple, top center frame at location [244, 0, 384, 137]
| [
244,
0,
384,
137
] | Detected green apple at location [244, 0, 384, 137] | Detected Green apple at location [389, 0, 575, 179]
Detected Green apple at location [245, 0, 385, 135]
Detected green apple at location [244, 0, 384, 137] | [
244,
0,
384,
137
] |
COCO_train2014_000000568848.jpg | refcoco_images/COCO_train2014_000000568848.jpg | food on left bun | [103.68,108.9,,332.45,303.4,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Food', 'Bun'] | Detected Food at location [101, 20, 509, 353]
Detected Bun at location [114, 107, 329, 296]
Detected food on left bun at location [103, 108, 329, 305]
| [
103,
108,
329,
305
] | Detected food at location [103, 108, 329, 305] | Detected Food at location [101, 20, 509, 353]
Detected Bun at location [114, 107, 329, 296]
Detected food at location [103, 108, 329, 305] | [
103,
108,
329,
305
] |
COCO_train2014_000000566495.jpg | refcoco_images/COCO_train2014_000000566495.jpg | right horse | [274.08,261.09,428.,,,575.35] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Horse'] | Detected Horse at location [0, 233, 273, 638]
Detected Horse at location [0, 234, 273, 563]
Detected right horse at location [274, 261, 427, 572]
| [
274,
261,
427,
572
] | Detected horse at location [274, 261, 427, 572] | Detected Horse at location [0, 233, 273, 638]
Detected Horse at location [0, 234, 273, 563]
Detected horse at location [274, 261, 427, 572] | [
274,
261,
427,
572
] |
COCO_train2014_000000564677.jpg | refcoco_images/COCO_train2014_000000564677.jpg | child in black and white sweater on left | [,,0.,,,,15.82,159.64,350.92] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Child', 'Sweater'] | Detected Child at location [158, 67, 410, 437]
Detected sweater at location [159, 247, 408, 437]
Detected child in black and white sweater on left at location [0, 18, 156, 359]
| [
0,
18,
156,
359
] | Detected child in black and white sweater at location [0, 18, 156, 359] | Detected Child at location [158, 67, 410, 437]
Detected sweater at location [159, 247, 408, 437]
Detected child in black and white sweater at location [0, 18, 156, 359] | [
0,
18,
156,
359
] |
COCO_train2014_000000537390.jpg | refcoco_images/COCO_train2014_000000537390.jpg | dark brown horse | [318.53,113.23,627.85,337.86] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Horse'] | Detected Horse at location [17, 107, 438, 402]
Detected dark brown horse at location [319, 113, 626, 338]
| [
319,
113,
626,
338
] | Detected dark brown horse at location [319, 113, 626, 338] | Detected Horse at location [17, 107, 438, 402]
Detected dark brown horse at location [319, 113, 626, 338] | [
319,
113,
626,
338
] |
COCO_train2014_000000581249.jpg | refcoco_images/COCO_train2014_000000581249.jpg | cat on suit | [227.5,,,68.44,480.,,,216.24] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Cat', 'Suit'] | Detected Cat at location [228, 64, 479, 213]
Detected cat on suit at location [226, 69, 479, 217]
| [
226,
69,
479,
217
] | Detected cat at location [226, 69, 479, 217] | Detected Cat at location [228, 64, 479, 213]
Detected cat at location [226, 69, 479, 217] | [
226,
69,
479,
217
] |
COCO_train2014_000000563601.jpg | refcoco_images/COCO_train2014_000000563601.jpg | blue box | [,18.87,183.96,411.55,420.98] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Blue box'] | Detected Blue box at location [14, 187, 414, 440]
Detected blue box at location [18, 184, 410, 420]
| [
18,
184,
410,
420
] | Detected blue box at location [18, 184, 410, 420] | Detected Blue box at location [14, 187, 414, 440]
Detected blue box at location [18, 184, 410, 420] | [
18,
184,
410,
420
] |
COCO_train2014_000000517148.jpg | refcoco_images/COCO_train2014_000000517148.jpg | large sheep on the right | [244.65,170.63,546.67,514.87] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Sheep'] | Detected Sheep at location [102, 163, 553, 556]
Detected Sheep at location [244, 167, 549, 521]
Detected large sheep on the right at location [245, 171, 546, 514]
| [
245,
171,
546,
514
] | Detected large sheep at location [245, 171, 546, 514] | Detected Sheep at location [102, 163, 553, 556]
Detected Sheep at location [244, 167, 549, 521]
Detected large sheep at location [245, 171, 546, 514] | [
245,
171,
546,
514
] |
COCO_train2014_000000538149.jpg | refcoco_images/COCO_train2014_000000538149.jpg | leftest one | [,11.51,,15.2,,205.54,306.25] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['One'] | Detected One at location [299, 136, 467, 439]
Detected leftest one at location [12, 16, 204, 306]
| [
12,
16,
204,
306
] | Detected leftest one at location [12, 16, 204, 306] | Detected One at location [299, 136, 467, 439]
Detected leftest one at location [12, 16, 204, 306] | [
12,
16,
204,
306
] |
COCO_train2014_000000527365.jpg | refcoco_images/COCO_train2014_000000527365.jpg | green stuff right | [383.7,,,73.52,640.,,,234.65] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Green stuff'] | Detected Green stuff at location [0, 42, 639, 422]
Detected Green stuff at location [0, 194, 257, 423]
Detected green stuff right at location [384, 75, 639, 234]
| [
384,
75,
639,
234
] | Detected green stuff at location [384, 75, 639, 234] | Detected Green stuff at location [0, 42, 639, 422]
Detected Green stuff at location [0, 194, 257, 423]
Detected green stuff at location [384, 75, 639, 234] | [
384,
75,
639,
234
] |
COCO_train2014_000000570346.jpg | refcoco_images/COCO_train2014_000000570346.jpg | elephant ear behind to left | [,64.43,156.28,402.,,,404.4,] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Elephant ear'] | Detected Elephant ear at location [119, 45, 500, 423]
Detected elephant ear behind to left at location [64, 157, 400, 405]
| [
64,
157,
400,
405
] | Detected elephant ear at location [64, 157, 400, 405] | Detected Elephant ear at location [119, 45, 500, 423]
Detected elephant ear at location [64, 157, 400, 405] | [
64,
157,
400,
405
] |
COCO_train2014_000000528157.jpg | refcoco_images/COCO_train2014_000000528157.jpg | boy in white and blue striped shirt | [235.15,110.77,424.1,,328.33] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Boy', 'White striped shirt', 'Blue striped shirt', 'Shirt'] | Detected Boy at location [297, 141, 549, 428]
Detected shirt at location [429, 219, 549, 373]
Detected shirt at location [429, 219, 549, 373]
Detected Shirt at location [476, 275, 639, 428]
Detected boy in white and blue striped shirt at location [298, 144, 547, 428]
| [
298,
144,
547,
428
] | Detected boy at location [298, 144, 547, 428] | Detected Boy at location [297, 141, 549, 428]
Detected shirt at location [429, 219, 549, 373]
Detected shirt at location [429, 219, 549, 373]
Detected Shirt at location [476, 275, 639, 428]
Detected boy at location [298, 144, 547, 428] | [
298,
144,
547,
428
] |
COCO_train2014_000000549691.jpg | refcoco_images/COCO_train2014_000000549691.jpg | far left goblet in front | [,,0.,,,129.88,153.6,,362.54] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Goblet'] | Detected Goblet at location [503, 128, 637, 377]
Detected Goblet at location [24, 126, 160, 366]
Detected far left goblet in front at location [26, 129, 157, 365]
| [
26,
129,
157,
365
] | Detected goblet at location [26, 129, 157, 365] | Detected Goblet at location [503, 128, 637, 377]
Detected Goblet at location [24, 126, 160, 366]
Detected goblet at location [26, 129, 157, 365] | [
26,
129,
157,
365
] |
COCO_train2014_000000553758.jpg | refcoco_images/COCO_train2014_000000553758.jpg | Boy on left | [172.89,103.29,307.11,460.87] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Boy'] | Detected Boy at location [171, 106, 306, 461]
Detected Boy on left at location [172, 103, 307, 461]
| [
172,
103,
307,
461
] | Detected Boy at location [172, 103, 307, 461] | Detected Boy at location [171, 106, 306, 461]
Detected Boy at location [172, 103, 307, 461] | [
172,
103,
307,
461
] |
COCO_train2014_000000580741.jpg | refcoco_images/COCO_train2014_000000580741.jpg | couch on left | [,,0.,,,143.53,297.11,335.86] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Couch'] | Detected Couch at location [2, 140, 297, 334]
Detected couch on left at location [0, 143, 296, 334]
| [
0,
143,
296,
334
] | Detected couch at location [0, 143, 296, 334] | Detected Couch at location [2, 140, 297, 334]
Detected couch at location [0, 143, 296, 334] | [
0,
143,
296,
334
] |
COCO_train2014_000000530394.jpg | refcoco_images/COCO_train2014_000000530394.jpg | left cat | [,,0.,,,180.3,,249.12,546.41] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Cat'] | Detected Cat at location [3, 183, 249, 547]
Detected left cat at location [0, 179, 250, 546]
| [
0,
179,
250,
546
] | Detected cat at location [0, 179, 250, 546] | Detected Cat at location [3, 183, 249, 547]
Detected cat at location [0, 179, 250, 546] | [
0,
179,
250,
546
] |
COCO_train2014_000000579186.jpg | refcoco_images/COCO_train2014_000000579186.jpg | right sofa | [302.57,224.75,558.14,360.79] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Sofa'] | Detected Sofa at location [302, 224, 560, 362]
Detected right sofa at location [303, 225, 557, 360]
| [
303,
225,
557,
360
] | Detected sofa at location [303, 225, 557, 360] | Detected Sofa at location [302, 224, 560, 362]
Detected sofa at location [303, 225, 557, 360] | [
303,
225,
557,
360
] |
COCO_train2014_000000522316.jpg | refcoco_images/COCO_train2014_000000522316.jpg | beige vase | [250.09,314.26,369.64,463.23] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Vase'] | Detected Vase at location [2, 265, 172, 511]
Detected beige vase at location [250, 315, 369, 462]
| [
250,
315,
369,
462
] | Detected beige vase at location [250, 315, 369, 462] | Detected Vase at location [2, 265, 172, 511]
Detected beige vase at location [250, 315, 369, 462] | [
250,
315,
369,
462
] |
COCO_train2014_000000546975.jpg | refcoco_images/COCO_train2014_000000546975.jpg | right toilet | [420.22,272.14,638.6,,371.13] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Toilet'] | Detected Toilet at location [2, 181, 156, 352]
Detected right toilet at location [420, 272, 636, 370]
| [
420,
272,
636,
370
] | Detected toilet at location [420, 272, 636, 370] | Detected Toilet at location [2, 181, 156, 352]
Detected toilet at location [420, 272, 636, 370] | [
420,
272,
636,
370
] |
COCO_train2014_000000544783.jpg | refcoco_images/COCO_train2014_000000544783.jpg | right duck | [287.57,,57.13,492.84,266.27] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Duck'] | Detected Duck at location [287, 53, 492, 269]
Detected Duck at location [121, 152, 314, 372]
Detected right duck at location [288, 57, 491, 265]
| [
288,
57,
491,
265
] | Detected duck at location [288, 57, 491, 265] | Detected Duck at location [287, 53, 492, 269]
Detected Duck at location [121, 152, 314, 372]
Detected duck at location [288, 57, 491, 265] | [
288,
57,
491,
265
] |
COCO_train2014_000000546148.jpg | refcoco_images/COCO_train2014_000000546148.jpg | hand | [,,0.81,,,0.81,574.71,311.57] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Hand'] | Detected Hand at location [40, 2, 576, 359]
Detected Hand at location [41, 2, 576, 270]
Detected hand at location [0, 0, 574, 310]
| [
0,
0,
574,
310
] | Detected hand at location [0, 0, 574, 310] | Detected Hand at location [40, 2, 576, 359]
Detected Hand at location [41, 2, 576, 270]
Detected hand at location [0, 0, 574, 310] | [
0,
0,
574,
310
] |
COCO_train2014_000000552116.jpg | refcoco_images/COCO_train2014_000000552116.jpg | man on ground blue | [188.57,181.39,524.55,305.83] | <|im_start|>system
You are a coordinate selection system. Your task is to:
1. Parse the given bounding box coordinates
2. For left/right selection:
- Compare the starting x-coordinates (first number in each box)
- Left means smaller x-coordinate
- Right means larger x-coordinate
3. For top/bottom selection:
... | ['Man', 'Ground', 'Blue'] | Detected Man at location [188, 180, 523, 310]
Detected man on ground blue at location [188, 182, 523, 306]
| [
188,
182,
523,
306
] | Detected man blue at location [188, 182, 523, 306] | Detected Man at location [188, 180, 523, 310]
Detected man blue at location [188, 182, 523, 306] | [
188,
182,
523,
306
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.