Datasets:
image image | objects dict | segmentation list |
|---|---|---|
{
"bbox": [
[
250,
119,
6,
10
],
[
139,
114,
11,
15
],
[
2,
0,
596,
136
],
[
124,
300,
172,
163
],
[
331,
105,
207,
325
],
[
267,
90,
31... | [
{
"label": "wall",
"category": 0,
"points": [
[
252,
119
],
[
250,
129
],
[
256,
129
]
]
},
{
"label": "wall",
"category": 0,
"points": [
[
144,
114
],
[
143,... | |
{
"bbox": [
[
0,
0,
349,
104
],
[
186,
231,
162,
14
],
[
291,
160,
17,
7
],
[
254,
147,
21,
19
],
[
75,
140,
35,
14
],
[
0,
136,
163,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
0,
0
],
[
0,
77
],
[
43,
78
],
[
51,
81
],
[
66,
81
],
[
68,
78
],
[
66,
... | |
{
"bbox": [
[
78,
1,
619,
420
],
[
753,
0,
13,
509
],
[
0,
376,
362,
134
],
[
371,
363,
390,
147
],
[
111,
0,
552,
73
],
[
277,
399,
84,... | [
{
"label": "wall",
"category": 0,
"points": [
[
78,
1
],
[
96,
11
],
[
94,
31
],
[
97,
42
],
[
101,
121
],
[
101,
314
],
[
... | |
{
"bbox": [
[
188,
97,
17,
15
],
[
162,
94,
12,
13
],
[
0,
1,
193,
146
],
[
0,
147,
9,
11
],
[
0,
100,
255,
154
],
[
191,
1,
63,
1... | [
{
"label": "wall",
"category": 0,
"points": [
[
202,
97
],
[
190,
98
],
[
188,
105
],
[
192,
109
],
[
192,
112
],
[
197,
110
],
[
... | |
{
"bbox": [
[
248,
128,
6,
18
],
[
246,
88,
8,
38
],
[
0,
2,
73,
129
],
[
75,
0,
180,
143
],
[
85,
225,
12,
17
],
[
107,
211,
13,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
254,
128
],
[
253,
133
],
[
248,
140
],
[
249,
144
],
[
254,
146
]
]
},
{
"label": "wall",
"category... | |
{
"bbox": [
[
31,
178,
132,
159
],
[
0,
73,
380,
404
],
[
289,
15,
478,
429
],
[
0,
322,
296,
189
],
[
31,
142,
133,
48
],
[
0,
0,
767,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
163,
187
],
[
81,
178
],
[
81,
190
],
[
79,
193
],
[
54,
190
],
[
31,
191
],
[
... | |
{
"bbox": [
[
73,
116,
14,
12
],
[
232,
1,
74,
7
],
[
112,
1,
123,
137
],
[
56,
1,
18,
92
],
[
1,
1,
43,
134
],
[
0,
157,
307,
92
... | [
{
"label": "wall",
"category": 0,
"points": [
[
73,
116
],
[
75,
128
],
[
87,
126
]
]
},
{
"label": "wall",
"category": 0,
"points": [
[
232,
8
],
[
245,
... | |
{
"bbox": [
[
396,
565,
114,
125
],
[
259,
456,
22,
3
],
[
74,
360,
161,
241
],
[
84,
1,
77,
5
],
[
0,
1,
51,
334
],
[
189,
0,
320,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
410,
565
],
[
396,
682
],
[
405,
690
],
[
412,
690
],
[
510,
669
],
[
509,
579
],
... | |
{
"bbox": [
[
0,
1,
681,
183
],
[
0,
227,
681,
283
],
[
229,
171,
283,
55
],
[
2,
227,
343,
35
],
[
27,
249,
21,
9
],
[
57,
245,
13,
... | [
{
"label": "sky",
"category": 2,
"points": [
[
1,
1
],
[
0,
94
],
[
121,
134
],
[
132,
135
],
[
143,
129
],
[
189,
120
],
[
... | |
{
"bbox": [
[
567,
314,
14,
43
],
[
1,
164,
77,
166
],
[
388,
28,
294,
330
],
[
85,
1,
321,
353
],
[
18,
1,
77,
13
],
[
155,
166,
120,
... | [
{
"label": "building",
"category": 1,
"points": [
[
580,
314
],
[
574,
324
],
[
572,
337
],
[
568,
341
],
[
570,
346
],
[
567,
352
],
... | |
{
"bbox": [
[
589,
247,
37,
30
],
[
2,
0,
395,
218
],
[
288,
1,
271,
64
],
[
351,
1,
331,
315
],
[
127,
359,
513,
152
],
[
1,
462,
219,
... | [
{
"label": "building",
"category": 1,
"points": [
[
594,
247
],
[
594,
266
],
[
589,
269
],
[
593,
274
],
[
598,
277
],
[
626,
276
],
... | |
{
"bbox": [
[
0,
0,
255,
170
],
[
0,
161,
255,
81
],
[
0,
235,
254,
20
]
],
"categories": [
2,
4,
13
],
"category_names": [
"sky",
"tree",
"earth"
]
} | [
{
"label": "sky",
"category": 2,
"points": [
[
0,
0
],
[
0,
169
],
[
6,
170
],
[
7,
166
],
[
15,
167
],
[
16,
159
],
[
25,... | |
{
"bbox": [
[
2,
1,
253,
112
],
[
42,
153,
18,
27
],
[
223,
88,
32,
13
],
[
0,
196,
80,
59
],
[
2,
148,
253,
107
],
[
2,
87,
253,
... | [
{
"label": "sky",
"category": 2,
"points": [
[
2,
1
],
[
2,
110
],
[
8,
113
],
[
21,
103
],
[
28,
111
],
[
53,
108
],
[
80... | |
{
"bbox": [
[
0,
0,
296,
190
],
[
164,
186,
132,
61
],
[
56,
1,
237,
22
],
[
2,
151,
182,
96
],
[
220,
144,
68,
70
],
[
33,
100,
42,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
296,
0
],
[
244,
18
],
[
57,
24
],
[
55,
21
],
[
55,
2
],
[
0,
1
],
[
1,
... | |
{
"bbox": [
[
265,
69,
2,
22
],
[
1,
37,
41,
28
],
[
20,
0,
234,
114
],
[
259,
219,
18,
6
],
[
260,
166,
13,
47
],
[
209,
0,
110,
... | [
{
"label": "building",
"category": 1,
"points": [
[
267,
69
],
[
265,
69
],
[
265,
91
]
]
},
{
"label": "building",
"category": 1,
"points": [
[
1,
37
],
[
2,... | |
{
"bbox": [
[
104,
285,
13,
18
],
[
376,
263,
6,
21
],
[
659,
167,
21,
111
],
[
573,
125,
82,
159
],
[
419,
100,
132,
182
],
[
1,
39,
42... | [
{
"label": "building",
"category": 1,
"points": [
[
104,
285
],
[
109,
294
],
[
115,
298
],
[
117,
303
],
[
116,
297
]
]
},
{
"label": "building",
"... | |
{
"bbox": [
[
64,
175,
30,
9
],
[
0,
0,
299,
223
],
[
3,
218,
24,
6
],
[
108,
0,
191,
54
],
[
23,
142,
276,
82
],
[
258,
58,
21,
8... | [
{
"label": "wall",
"category": 0,
"points": [
[
94,
175
],
[
88,
175
],
[
64,
181
],
[
64,
184
]
]
},
{
"label": "wall",
"category": 0,
"points": [
[
0,
... | |
{
"bbox": [
[
27,
167,
16,
7
],
[
211,
25,
92,
126
],
[
0,
8,
209,
166
],
[
0,
219,
29,
9
],
[
0,
207,
37,
11
],
[
0,
175,
60,
31
... | [
{
"label": "wall",
"category": 0,
"points": [
[
27,
167
],
[
27,
174
],
[
42,
174
],
[
43,
167
]
]
},
{
"label": "wall",
"category": 0,
"points": [
[
211,
... | |
{
"bbox": [
[
41,
322,
28,
14
],
[
487,
300,
118,
50
],
[
610,
299,
24,
41
],
[
638,
297,
13,
6
],
[
657,
295,
25,
37
],
[
239,
289,
24,... | [
{
"label": "building",
"category": 1,
"points": [
[
41,
335
],
[
48,
336
],
[
64,
333
],
[
67,
331
],
[
69,
324
],
[
67,
322
],
[
... | |
{
"bbox": [
[
0,
20,
303,
423
],
[
0,
0,
304,
213
],
[
66,
350,
238,
94
]
],
"categories": [
1,
2,
4
],
"category_names": [
"building",
"sky",
"tree"
]
} | [
{
"label": "building",
"category": 1,
"points": [
[
76,
20
],
[
76,
26
],
[
79,
29
],
[
77,
31
],
[
72,
31
],
[
61,
37
],
[
... | |
{
"bbox": [
[
30,
17,
310,
453
],
[
2,
0,
371,
440
],
[
0,
386,
372,
111
]
],
"categories": [
1,
2,
4
],
"category_names": [
"building",
"sky",
"tree"
]
} | [
{
"label": "building",
"category": 1,
"points": [
[
54,
25
],
[
54,
39
],
[
39,
41
],
[
41,
57
],
[
33,
58
],
[
32,
70
],
[
... | |
{
"bbox": [
[
0,
0,
254,
190
],
[
0,
162,
254,
92
],
[
218,
155,
23,
13
],
[
199,
155,
17,
5
],
[
54,
40,
194,
198
],
[
246,
50,
8,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
0,
0
],
[
0,
190
],
[
1,
151
],
[
3,
149
],
[
10,
149
],
[
9,
145
],
[
5,
... | |
{
"bbox": [
[
1,
1,
508,
294
],
[
6,
405,
281,
3
],
[
259,
362,
25,
38
],
[
85,
269,
424,
138
],
[
151,
249,
2,
21
],
[
2,
229,
284,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
1,
1
],
[
1,
295
],
[
2,
228
],
[
83,
269
],
[
87,
268
],
[
22,
238
],
[
1... | |
{
"bbox": [
[
100,
57,
10,
420
],
[
53,
1,
241,
477
],
[
2,
1,
44,
477
],
[
357,
0,
254,
477
],
[
193,
281,
254,
197
],
[
50,
21,
51,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
100,
57
],
[
103,
468
],
[
104,
477
],
[
110,
477
],
[
106,
59
]
]
},
{
"label": "wall",
"category":... | |
{
"bbox": [
[
125,
17,
17,
6
],
[
0,
13,
355,
241
],
[
354,
0,
2,
23
],
[
178,
0,
130,
24
],
[
153,
0,
26,
15
],
[
105,
0,
38,
16
... | [
{
"label": "building",
"category": 1,
"points": [
[
142,
23
],
[
142,
17
],
[
125,
18
],
[
135,
19
]
]
},
{
"label": "building",
"category": 1,
"points": [
[
... | |
{
"bbox": [
[
109,
148,
60,
77
],
[
184,
139,
53,
114
],
[
2,
1,
253,
224
],
[
2,
88,
253,
167
],
[
23,
8,
89,
216
]
],
"categories": [
1,
1,
... | [
{
"label": "building",
"category": 1,
"points": [
[
111,
149
],
[
109,
193
],
[
124,
198
],
[
118,
212
],
[
112,
218
],
[
112,
222
],
... | |
{
"bbox": [
[
378,
297,
132,
130
],
[
0,
0,
510,
327
],
[
0,
1,
489,
529
],
[
1,
410,
306,
268
],
[
320,
409,
190,
270
],
[
185,
408,
50... | [
{
"label": "building",
"category": 1,
"points": [
[
510,
297
],
[
433,
328
],
[
434,
332
],
[
423,
343
],
[
408,
347
],
[
404,
354
],
... | |
{
"bbox": [
[
385,
136,
13,
55
],
[
0,
0,
397,
152
],
[
255,
191,
143,
74
],
[
229,
1,
150,
23
],
[
125,
84,
54,
49
],
[
87,
72,
34,
... | [
{
"label": "wall",
"category": 0,
"points": [
[
398,
136
],
[
398,
159
],
[
385,
166
],
[
385,
190
],
[
398,
191
]
]
},
{
"label": "wall",
"category... | |
{
"bbox": [
[
340,
259,
22,
11
],
[
275,
171,
53,
95
],
[
1,
0,
509,
324
],
[
0,
254,
440,
237
],
[
364,
232,
14,
14
],
[
88,
97,
304,
... | [
{
"label": "sky",
"category": 2,
"points": [
[
340,
259
],
[
352,
270
],
[
362,
259
],
[
349,
263
]
]
},
{
"label": "sky",
"category": 2,
"points": [
[
281,
... | |
{
"bbox": [
[
195,
0,
235,
66
],
[
253,
31,
85,
34
],
[
0,
0,
207,
181
],
[
0,
167,
430,
100
],
[
336,
181,
25,
13
],
[
48,
18,
381,
... | [
{
"label": "sky",
"category": 2,
"points": [
[
195,
1
],
[
195,
5
],
[
206,
15
],
[
204,
20
],
[
209,
22
],
[
219,
17
],
[
... | |
{
"bbox": [
[
140,
383,
13,
9
],
[
473,
368,
29,
21
],
[
425,
361,
7,
12
],
[
624,
205,
57,
80
],
[
614,
97,
67,
120
],
[
0,
0,
613,
... | [
{
"label": "building",
"category": 1,
"points": [
[
153,
390
],
[
148,
383
],
[
140,
390
],
[
140,
392
],
[
153,
392
]
]
},
{
"label": "building",
"... |
ADE-20K
22,207 images and 396,285 annotations across 150 classes, annotated as polygon.
View on Pictograph · Pictograph Research · Custom License
About
ADE-20K is a computer-vision dataset curated and annotated on Pictograph. On Pictograph you can browse every annotated image, fork it into your own workspace in one click, export it in a dozen formats, or train a model on it directly.
At a glance
| Metric | Value |
|---|---|
| Images | 22,207 |
| Annotations | 396,285 |
| Classes | 150 |
| Annotation types | polygon |
| Splits | train |
Dataset structure
This dataset uses the Hugging Face imagefolder layout: each split directory holds the images plus a metadata.jsonl that links every image to its annotations by file_name.
| Field | Description |
|---|---|
file_name |
Path to the image within the split directory. |
objects.bbox |
Bounding boxes as [x, y, width, height] (pixels). |
objects.categories |
Integer class index per box (matches the class list below). |
objects.category_names |
Human class name per box. |
segmentation |
Polygon rings as [[x, y], ...] with class label. |
Use it
from datasets import load_dataset
ds = load_dataset("pictograph/ade-20k")
print(ds)
Prefer a full annotation editor, one-click fork, multi-format export, and one-click training? Open this dataset on Pictograph.
Classes
Class index matches objects.categories in metadata.jsonl.
All 150 classes, in index order: wall, building, sky, floor, tree, ceiling, road, bed, windowpane, grass, cabinet, sidewalk, person, earth, door, table, mountain, plant, curtain, chair, car, water, painting, sofa, shelf, house, sea, mirror, rug, field, armchair, seat, fence, desk, rock, wardrobe, lamp, bathtub, railing, cushion, base, box, column, signboard, chest_of_drawers, counter, sand, sink, skyscraper, fireplace, refrigerator, grandstand, path, stairs, runway, case, pool_table, pillow, screen_door, stairway, river, bridge, bookcase, blind, coffee_table, toilet, flower, book, hill, bench, countertop, stove, palm, kitchen_island, computer, swivel_chair, boat, bar, arcade_machine, hovel, bus, towel, light, truck, tower, chandelier, awning, streetlight, booth, television_receiver, airplane, dirt_track, apparel, pole, land, bannister, escalator, ottoman, bottle, buffet, poster, stage, van, ship, fountain, conveyer_belt, canopy, washer, plaything, swimming_pool, stool, barrel, basket, waterfall, tent, bag, minibike, cradle, oven, ball, food, step, tank, trade_name, microwave, pot, animal, bicycle, lake, dishwasher, screen, blanket, sculpture, hood, sconce, vase, traffic_light, tray, ashcan, fan, pier, crt_screen, plate, monitor, bulletin_board, shower, radiator, glass, clock, flag.
License
ADE20K uses a dual license. Images: non-commercial research and educational use only (bespoke MIT CSAIL agreement; access via request form; MIT does not hold image copyright; commercial use requires permission from MIT CSAIL). Annotations, software, and website: BSD-3-Clause, Copyright 2019 MIT, CSAIL.
Source and attribution
This dataset is derived from ADE20K, created by MIT CSAIL - Zhou, Zhao, Puig, Fidler, Barriuso, Torralba, originally licensed BSD 3-Clause License. We are grateful to the original authors. If you use this data, please cite the original source above.
Published from Pictograph - annotate, train, and deploy from one API.
- Downloads last month
- -