Datasets:
clip1_uid large_string | clip1_start_frame int64 | clip1_end_frame int64 | clip2_uid large_string | clip2_start_frame int64 | clip2_end_frame int64 | V large_string | ARG1 large_string | label int64 | video_uid large_string | start_frame int64 | end_frame int64 | split large_string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
0169d43e-483e-4d65-83a8-3f604b1dd724 | 8,764 | 9,004 | Not required | -1 | -1 | plays | lawn tennis | 2 | a843bc1d-8a58-4aea-9c56-b7c042011941 | 26,523 | 26,763 | train |
20d02685-499c-41f8-ba9f-4414332aef29 | 1,862 | 2,102 | Not required | -1 | -1 | uses | the laptop | 2 | 48f344b2-67e5-46ad-9eac-a995b24dcbc7 | 13,593 | 13,833 | train |
47ecc062-c675-4e59-ac55-76fad8ecafdc | 3,035 | 3,275 | Not required | -1 | -1 | adjusts | the napkin | 0 | 5efa966b-01b6-4bbb-a4b6-702ef89c3a10 | 20,794 | 21,034 | train |
f93e485e-9ec3-4856-809d-f050c6439b7b | 3,491 | 3,731 | Not required | -1 | -1 | hits | wood frame | 3 | 2b89ddf6-d019-4664-87a4-ea308717cd34 | 3,491 | 3,731 | train |
370456d4-864e-4349-9cd0-aa80d01fe695 | 8,984 | 9,224 | Not required | -1 | -1 | wipes | a wooden stand with the left hand | 1 | 380de10c-99d3-4270-b908-6765c388a116 | 17,743 | 17,983 | train |
e8db5f9a-d30b-438b-ae3a-f002cd5dd6de | 8,305 | 8,545 | Not required | -1 | -1 | picks | a paper | 1 | cb235c78-e524-417a-89fd-57b682bb223b | 17,064 | 17,304 | train |
bb9046a7-5cf3-4188-93b3-a6fcaf67add4 | 7,121 | 7,361 | Not required | -1 | -1 | plays | a guitar | 3 | 712ef11e-dc19-4fe1-ae66-0d9ec934196e | 96,880 | 97,120 | train |
d330e012-1b7e-43ab-b64f-d1386c0f759f | 7,921 | 8,161 | Not required | -1 | -1 | covers | the pan | 0 | 597b1ca1-e2ea-4a0e-8be0-31b713151ba2 | 52,680 | 52,920 | train |
3c18ec5c-7e6b-400d-9260-70598e04a373 | 7,402 | 7,642 | Not required | -1 | -1 | drops | the leaves | 3 | d9941963-10dd-49bc-badc-998c88cf1e18 | 34,161 | 34,401 | train |
ddf5516c-f225-4bc0-947d-da33b2435733 | 3,942 | 4,182 | Not required | -1 | -1 | cuts | the third branch | 1 | abc7e388-83f7-4bea-8bf2-09211e8c104b | 107,443 | 107,683 | train |
63b3c6aa-a327-4c86-8e9b-c5f51462dce4 | 3,377 | 3,617 | Not required | -1 | -1 | takes | the dish | 3 | 28313dbe-ad68-43cb-bb72-602d5600d15a | 3,377 | 3,617 | train |
8cc25698-9f3a-4dd2-a3bc-a0187f221477 | 603 | 843 | Not required | -1 | -1 | moves | the dough | 0 | 89857b33-fa50-469a-bbb3-91c8ab655931 | 603 | 843 | train |
c743e9d3-f5e6-4c9f-b006-f28062e62d3d | 6,269 | 6,509 | Not required | -1 | -1 | chops | another potato | 2 | 559ddb1f-f0c0-4d27-b2c2-ebabe103dc3b | 15,028 | 15,268 | train |
26bd83b9-4724-46a2-b98e-ee405c1088bd | 2,996 | 3,236 | Not required | -1 | -1 | picks | the plates | 3 | bcb6ca41-64ea-4c2b-90c7-1e523a5a324e | 2,996 | 3,236 | train |
303505b9-2cf3-488d-ae5f-d72d3bf73585 | 5,211 | 5,451 | Not required | -1 | -1 | places | her left index finger | 0 | 38802829-2f8b-4023-b00e-23c00570b820 | 40,971 | 41,211 | train |
5c684ceb-fca8-45ab-8deb-f2a2302fc961 | 1,927 | 2,167 | Not required | -1 | -1 | uses | the paint brush | 1 | 4aeb7f3c-855a-4c66-b3d9-02b0eb963883 | 10,686 | 10,926 | train |
80397012-d848-4872-a82c-bbeeb53ccaf1 | 1,785 | 2,025 | Not required | -1 | -1 | puts | the tap | 0 | edc1869c-8a97-44fd-ab47-63fda4a54df9 | 10,545 | 10,785 | train |
6974814e-cbf7-4248-9b20-0b09181c2f7b | 6,514 | 6,754 | Not required | -1 | -1 | picks | the seed | 0 | fa2f1291-3796-41a6-8f7b-6e7c1491b9b2 | 6,514 | 6,754 | train |
8a1e4ca7-acc4-450c-87f1-e58952d8452e | 7,791 | 8,031 | Not required | -1 | -1 | throws | mortar | 3 | eaa817af-b7e0-4f0a-8156-73834cf1157b | 16,550 | 16,790 | train |
ec18ebd3-3c61-4f4c-8616-a4b445cd28e5 | 8,776 | 9,016 | Not required | -1 | -1 | cuts | sandpaper | 2 | 91c006d8-b55c-4426-a7b6-dd527a8ec27d | 144,504 | 144,744 | train |
10861314-64d8-4774-925f-2ea0720fea69 | 7,472 | 7,712 | Not required | -1 | -1 | moves | a plate | 2 | 27e6c383-64b8-41c3-80ac-87c20d6e588b | 88,231 | 88,471 | train |
318dd659-4220-4dfa-b192-bc50745a7d06 | 5,775 | 6,015 | Not required | -1 | -1 | picks | wooden spoon | 0 | 105d3303-8e2d-4c20-96ff-e9a8ff325109 | 23,534 | 23,774 | train |
085ac5b9-eb2a-4f4e-929a-f20b93558803 | 2,977 | 3,217 | Not required | -1 | -1 | takes | two small wood pieces | 1 | 34af64b6-2135-43dc-9a33-2899d92a0c8e | 2,976 | 3,216 | train |
3ae0de07-c729-40c7-b7bb-01aa3d9f5818 | 1,981 | 2,221 | Not required | -1 | -1 | pulls | the leaves | 3 | 155f8d74-4c5c-4821-a18b-fceaa9c6199c | 42,482 | 42,722 | train |
4037d766-ae81-4ba8-841d-a7dde93f724e | 1,014 | 1,254 | Not required | -1 | -1 | picks | a paper | 2 | 8e58a7b3-43ef-406d-ad5d-901f83418261 | 198,773 | 199,013 | train |
e658e265-75b4-4f98-bb51-aa82dd543d4a | 9,043 | 9,283 | Not required | -1 | -1 | picks | dry weeds | 2 | 2fd1837a-613b-48af-9ad2-0222f8fd6b69 | 26,802 | 27,042 | train |
6cd59ad6-672c-488b-894c-b7ffecab7799 | 5,287 | 5,527 | Not required | -1 | -1 | moves | phone | 0 | b75bf090-1439-4b11-862b-d1dadab7f854 | 77,047 | 77,287 | train |
1560b9ee-e2da-4dda-a875-52e5bae8d88a | 6,416 | 6,656 | Not required | -1 | -1 | throws | the excess mortar | 3 | 58b068dd-758f-43c7-8cd5-d17cc1aafb5a | 6,415 | 6,655 | train |
6cac6b65-0c6d-48b8-aee1-21f648329dc8 | 5,202 | 5,442 | Not required | -1 | -1 | checks | on a pile of materials | 1 | dc496391-e201-4f7a-a2b9-1aca69a171e7 | 77,446 | 77,686 | train |
4713085d-6f54-42f6-a4a7-1e0cdf438df8 | 4,483 | 4,723 | Not required | -1 | -1 | dips | the paint brush in his right hand | 0 | aa7bf4c1-0482-42c0-9d78-088870225045 | 22,242 | 22,482 | train |
9143d3af-3626-413b-b547-45aac83d7067 | 2,730 | 2,970 | Not required | -1 | -1 | moves | a bag | 3 | 4abd8edc-4751-4a47-9808-696d960b7557 | 119,974 | 120,214 | train |
09e3904d-cc9c-441e-ae53-5db8940a1890 | 8,618 | 8,858 | Not required | -1 | -1 | picks | a knife | 2 | 8996ece6-f2c1-49c4-aab7-fee6c30f2ca2 | 17,377 | 17,617 | train |
6f4ab58b-7709-4cc2-9eec-b1f16ec771a0 | 6,201 | 6,441 | Not required | -1 | -1 | holds | chocolate | 0 | a6fb31a3-eca4-4e8f-a23b-16e4f2a9269b | 104,960 | 105,200 | train |
69a4857f-0af9-4f89-b7bd-372fac01a11a | 141 | 381 | Not required | -1 | -1 | separates | the cut out paper on the table | 3 | f4c4c22e-2719-4897-87a8-bb754903eb01 | 2,967 | 3,207 | train |
2ff8c772-c1d6-464a-84c8-dcc1532fc5a1 | 6,855 | 7,095 | Not required | -1 | -1 | dips | the brush | 0 | 4b63760e-3016-4a13-ad93-c9487a433a4c | 33,615 | 33,855 | train |
e7994ab9-49b2-473b-b16b-eca6786b5a59 | 4,650 | 4,890 | Not required | -1 | -1 | holds | a spanner | 2 | 0e6fb738-05fc-4dd5-9746-a8e10efe8c20 | 4,650 | 4,890 | train |
148ea873-1cd5-4f74-a9ce-751ce6513367 | 1,445 | 1,685 | Not required | -1 | -1 | checks | the smoothness of the railing | 0 | 2a00e878-bd87-4d39-91c4-f27fcb7b5feb | 1,445 | 1,685 | train |
d35b3296-0ae5-421a-b6c3-1612e5be8cf1 | 1,251 | 1,491 | Not required | -1 | -1 | rubs | sand | 1 | f786bdd6-2d59-432d-8c2e-84482e2032c5 | 145,979 | 146,219 | train |
d768ffc1-8225-4a63-9d5a-0b0a2bda02a0 | 8,669 | 8,909 | Not required | -1 | -1 | drops | cheese particle | 1 | 2409a5a7-a4ed-4fcb-ad77-024dc20988ca | 8,668 | 8,908 | train |
ae172aa5-eb02-448f-8b80-20d213a46991 | 7,285 | 7,525 | Not required | -1 | -1 | plays | the card them | 2 | c48a70f7-44a3-44aa-ac14-baf35e696e5c | 34,044 | 34,284 | train |
cfae9bf1-dcee-4315-9070-8a32605b4124 | 0 | 217 | Not required | -1 | -1 | drops | the tapping block | 1 | 2353f031-31de-4d26-b639-474ea59a39f0 | 0 | 217 | train |
484ac767-91bf-40c2-86b2-d03de133074c | 4,447 | 4,687 | Not required | -1 | -1 | paints | the wardrobe | 3 | 150c7fa0-941b-4565-aff6-e83c7b6daf31 | 67,206 | 67,446 | train |
ec2e69c1-fd07-48ec-adff-0b2cf3ab25b6 | 0 | 128 | Not required | -1 | -1 | picks | a weight plate | 1 | 2c84e6a0-2a0b-4de8-ae99-aedbc871dffa | 0 | 128 | train |
c6cea8fc-7f7c-434b-9d28-1af63b601ec9 | 412 | 652 | Not required | -1 | -1 | rubs | oil on the motorcycle | 1 | 1e64cbac-80af-4aa8-86bd-cc03c081ab1a | 411 | 651 | train |
3ecc5e01-ae26-46b5-80bf-a4a7ac29029d | 8,694 | 8,934 | Not required | -1 | -1 | holds | the ceiling plate bracket | 3 | 8da5f5a1-ae5f-45e8-a7f7-3226547c3c4d | 26,453 | 26,693 | train |
a1711833-58f2-4fd0-93e2-4bb5c7584d34 | 769 | 1,009 | Not required | -1 | -1 | drops | the towel | 3 | 52594838-2533-465c-8be4-2d86e9fe5ef1 | 9,529 | 9,769 | train |
a3033be0-026f-44d4-92f0-033c32d25e3e | 5,215 | 5,455 | Not required | -1 | -1 | turns | the mold container | 0 | f786bdd6-2d59-432d-8c2e-84482e2032c5 | 5,215 | 5,455 | train |
28de2fac-714a-4322-ad26-0ad36fb26f5f | 3,685 | 3,925 | Not required | -1 | -1 | puts | food | 0 | be8889c4-114f-4cb2-9e2c-fef576dbb00d | 48,444 | 48,684 | train |
8bcd34ce-09f1-4845-9c38-c56e0ddaf969 | 1,595 | 1,835 | Not required | -1 | -1 | takes | chicken | 1 | e4ad6fd7-2e3e-4991-b392-a0056f702286 | 69,096 | 69,336 | train |
d525b1b4-e489-4717-ba96-995403e1a7c1 | 3,240 | 3,480 | Not required | -1 | -1 | rolls | the clay | 3 | e6d01674-2031-4073-8b5c-adef89cd96d1 | 3,240 | 3,480 | train |
d21af9c2-6172-493a-9543-f1f92325aae7 | 7,796 | 8,036 | Not required | -1 | -1 | drops | his left hand | 0 | d3a0899e-2093-454c-9f65-30087883193a | 52,556 | 52,796 | train |
40ac21f0-722f-411d-b920-625a3a094703 | 5,882 | 6,122 | Not required | -1 | -1 | raises | the left leg | 2 | a95506d4-a846-42f7-a999-e77d543940c2 | 5,882 | 6,122 | train |
34262ef4-0094-4cf8-a371-66262858ab77 | 7,812 | 8,052 | Not required | -1 | -1 | rotates | a cable | 0 | 5f2be256-1298-4876-a310-cc9c7e80774a | 16,571 | 16,811 | train |
387f9c84-bb32-4265-ad89-43cfd3ec3292 | 7,283 | 7,523 | Not required | -1 | -1 | drops | the hose | 3 | 4f52acb9-9f42-4424-a4ac-2d3bc8fd5a5f | 16,043 | 16,283 | train |
b8d99b61-7d37-4303-88c2-8dff488f10ef | 5,542 | 5,782 | Not required | -1 | -1 | takes | his left hand | 3 | d78e3794-6f26-4e01-a2ce-08696774c056 | 23,301 | 23,541 | train |
1e1e832e-8662-402b-adcd-c524ef62bb25 | 238 | 478 | Not required | -1 | -1 | closes | the dry cleaning machine | 0 | bda3cdb5-32ef-433f-bc6a-e77617447d30 | 238 | 478 | train |
36f8b38a-a408-48c2-863f-26397fd45bbf | 4,722 | 4,962 | Not required | -1 | -1 | brings | nylon bag | 3 | 114d86a7-2849-46de-8bb7-8fe1e1a48be8 | 4,721 | 4,961 | train |
a37c1aed-a0e6-4f0d-8748-92ea6d534f53 | 1,677 | 1,917 | Not required | -1 | -1 | shakes | water | 2 | 4453ce6d-e2f9-45eb-8694-daf9057daacf | 1,676 | 1,916 | train |
833a377f-ffd7-4e18-8c26-78ce56abb099 | 2,082 | 2,322 | Not required | -1 | -1 | picks | an egg | 0 | 68b90b42-7e3c-443f-9bac-e1fb46ee40b8 | 2,082 | 2,322 | train |
381e7ae9-2eae-4534-8df8-2e7793e8c5e9 | 7,031 | 7,271 | Not required | -1 | -1 | trims | the plastic conduit | 1 | ae7b6096-4f00-42af-857d-603c2cbfa940 | 78,790 | 79,030 | train |
2c3d9cd1-e935-4023-bb4c-c957173e0743 | 8,476 | 8,716 | Not required | -1 | -1 | holds | the plier | 0 | af73617b-ee77-4255-baf3-514508c62353 | 26,236 | 26,476 | train |
0ea0f6cf-326a-4fc6-bd17-bcdb1c2eb5c7 | 2,043 | 2,283 | Not required | -1 | -1 | moves | the plank on the floor | 3 | 18d704a0-736d-4423-9d1e-ceaea3423d93 | 64,802 | 65,042 | train |
3122b423-0fd4-46a9-94f0-6fd14a333b0a | 8,817 | 9,057 | Not required | -1 | -1 | picks | a napkin | 0 | 7021640c-5533-49e1-b2e1-e638eb6bb2c9 | 44,576 | 44,816 | train |
e127fc34-0de5-41b0-ab68-7d5574bcf613 | 1,250 | 1,490 | Not required | -1 | -1 | pulls | a book | 2 | 62ca20e4-b289-47fc-b175-4ce77178de82 | 1,249 | 1,489 | train |
0e3c614c-67e8-4802-80ce-69b5f78c6498 | 2,775 | 3,015 | Not required | -1 | -1 | holds | tripod stand | 1 | 81be6ac3-8fe1-49c1-849f-06aebada2849 | 11,535 | 11,775 | train |
6205c5fe-9f25-49ac-95d0-61834d2d5ce2 | 2,689 | 2,929 | Not required | -1 | -1 | drops | the ball | 1 | 98fdad00-49db-478f-bc77-c7d06992882e | 29,448 | 29,688 | train |
ad25b4d9-2899-49e8-8d08-f080853b18be | 688 | 928 | Not required | -1 | -1 | drives | a nail | 3 | e406c375-245c-419e-9525-652f61eda7d3 | 63,447 | 63,687 | train |
80264f3a-b986-4189-80af-f1dcbaa0fc14 | 3,759 | 3,946 | Not required | -1 | -1 | removes | a rubber | 1 | 9da4b8a6-d09f-433e-ab23-96032e2b7aa7 | 3,759 | 3,946 | train |
e80a8a2d-7947-4aa2-9cf2-98b606a467b7 | 1,518 | 1,758 | Not required | -1 | -1 | scoops | the stew | 0 | 094bb63c-2050-4471-88eb-de3b86c26c81 | 19,277 | 19,517 | train |
91b689b4-b4c4-4051-999e-1d4ea529ebba | 7,560 | 7,800 | Not required | -1 | -1 | moves | the fluorescent light | 1 | 4e3fc1e9-424f-4921-9068-d468c135f347 | 7,560 | 7,800 | train |
3cc7a40c-2990-4960-9871-be71b7386e69 | 5,836 | 6,076 | Not required | -1 | -1 | lifts | a bowl | 3 | a77c0e56-3880-48bf-b6bf-9d46c6a42fc7 | 299,148 | 299,388 | train |
d9318bc0-7fa6-4207-83f6-3457f6fe40be | 472 | 712 | Not required | -1 | -1 | takes | a spray | 1 | f5c456b2-b998-4f42-82bd-786833fb3891 | 18,231 | 18,471 | train |
ababf69e-2711-4c64-890b-1a345aec2311 | 4,015 | 4,255 | Not required | -1 | -1 | throws | the tennis ball | 3 | 712ef11e-dc19-4fe1-ae66-0d9ec934196e | 75,774 | 76,014 | train |
4579b8ec-7e66-4437-8059-d95aaaaf8cc1 | 560 | 800 | Not required | -1 | -1 | drops | the mortar | 0 | c003438a-eba5-430e-9f54-95a4d568e511 | 117,804 | 118,044 | train |
661e7a9b-2d4a-42f7-b6c0-8128fa01dac8 | 3,247 | 3,487 | Not required | -1 | -1 | puts | the scissors | 1 | 73803873-303f-484c-b647-0b6dd8f6c1c3 | 30,006 | 30,246 | train |
41b8254c-ca1e-464c-9323-55301fb5f0e8 | 563 | 803 | Not required | -1 | -1 | drops | the spanner | 2 | ba11fcda-0048-4440-a7e8-fd15d1661a27 | 562 | 802 | train |
ea87324e-d129-425f-b247-e6bcc4ff332c | 2,233 | 2,473 | Not required | -1 | -1 | dips | the paint brush | 0 | 4aeb7f3c-855a-4c66-b3d9-02b0eb963883 | 2,232 | 2,472 | train |
dade9249-30bf-4f55-873f-b38837f4e4e1 | 7,409 | 7,649 | Not required | -1 | -1 | puts | the cloth | 2 | ae7b6096-4f00-42af-857d-603c2cbfa940 | 70,168 | 70,408 | train |
c3deacc6-9b6c-4250-bf08-466c8c1eaed3 | 0 | 120 | Not required | -1 | -1 | sets | the camera | 0 | 7ab911c5-44ec-44e8-b81a-e05625f39500 | 0 | 120 | train |
14a7eeef-3cb5-47f6-b984-dd2456eb799d | 2,145 | 2,385 | Not required | -1 | -1 | takes | a needle | 1 | 74736d57-1b15-4bad-9392-7b1b4ac39617 | 20,378 | 20,618 | train |
76f49696-c7cb-4aaf-83c2-0b58bc7ff9d2 | 960 | 1,200 | Not required | -1 | -1 | peels | dirt | 1 | 6eb083c4-a7dc-454b-84bc-2f0d8a69dffe | 960 | 1,200 | train |
d7d7a221-a475-440e-aeb2-f0c9530aa2c6 | 427 | 667 | Not required | -1 | -1 | unfolds | the napkin | 1 | 7021640c-5533-49e1-b2e1-e638eb6bb2c9 | 18,186 | 18,426 | train |
c2c8c5fa-3278-4d2e-b3f3-31bda3aab167 | 6,612 | 6,852 | Not required | -1 | -1 | drops | the paint brush | 1 | 0b22b43c-f57e-4f4f-8840-3ca0bf086b9c | 51,371 | 51,611 | train |
2029a01d-9e6f-4e2e-ab4f-9587d3d693e3 | 4,497 | 4,737 | Not required | -1 | -1 | shakes | strainer | 2 | 0fe191ef-c28a-422c-aede-46f8aa8532a6 | 94,256 | 94,496 | train |
6f082d5d-5f31-4358-b2dc-16d320312ab3 | 4,675 | 4,915 | Not required | -1 | -1 | arranges | the doughs | 0 | f938bcd9-bf30-4dfb-9a99-d6b9ee53c046 | 4,674 | 4,914 | train |
b84c1c9a-6c5f-47cd-8333-929bab273872 | 7,279 | 7,519 | Not required | -1 | -1 | takes | a pocket knife | 3 | 8da5f5a1-ae5f-45e8-a7f7-3226547c3c4d | 7,279 | 7,519 | train |
9143d3af-3626-413b-b547-45aac83d7067 | 7,177 | 7,417 | Not required | -1 | -1 | turns | the brick mould | 0 | 4abd8edc-4751-4a47-9808-696d960b7557 | 124,421 | 124,661 | train |
f7fe40c8-dede-4eaf-90a8-b23e98301ab1 | 8,825 | 9,065 | Not required | -1 | -1 | turns | the food | 1 | f7a0beb6-b220-40c0-a72d-ec4b79134a73 | 8,825 | 9,065 | train |
6b1fd478-dba5-45c9-b057-c66e92bb8b88 | 6,005 | 6,245 | Not required | -1 | -1 | pours | sand | 2 | 155f8d74-4c5c-4821-a18b-fceaa9c6199c | 242,794 | 243,034 | train |
9dd1aa17-0d02-444e-afbf-b00ba71fbd96 | 4,619 | 4,859 | Not required | -1 | -1 | opens | a kitchen shelf | 1 | f7f7d2ae-5d75-4447-934d-1573f95b8f81 | 13,378 | 13,618 | train |
17eb6f00-a1df-4d4a-b1ef-a5ade6ec4872 | 2,458 | 2,698 | Not required | -1 | -1 | drops | the small metal rod | 1 | c976bf0b-e005-40b6-8482-6c1431797edc | 92,702 | 92,942 | train |
b41e3f81-6457-4cd5-be6a-aa0bfccc07e7 | 675 | 915 | Not required | -1 | -1 | unfolds | the napkin | 1 | 7021640c-5533-49e1-b2e1-e638eb6bb2c9 | 45,434 | 45,674 | train |
084e02d1-1f57-463a-adf4-24ff45633444 | 3,046 | 3,286 | Not required | -1 | -1 | drills | another screw | 3 | a0d1444a-7f22-4575-adfe-d7a27436c545 | 3,046 | 3,286 | train |
d9f0fd9d-f44f-4295-b5ee-2b4fb70cb46d | 8,454 | 8,694 | Not required | -1 | -1 | looses | the nut | 1 | f522ce65-50a9-4119-9d49-57c32dea58f7 | 44,213 | 44,453 | train |
ee553438-45a5-47bb-82b7-f2f65f718ecb | 4,376 | 4,616 | Not required | -1 | -1 | removes | bolt | 2 | ffb7ecf6-f44e-499b-b315-a4aeabf3578c | 67,135 | 67,375 | train |
2396f2ce-f0d5-4232-9889-5a41a4c15b48 | 3,240 | 3,480 | Not required | -1 | -1 | places | the dish | 1 | 1fae6ecb-2ad9-4160-b388-c34e7d018915 | 65,999 | 66,239 | train |
59ef4ab3-4af5-4b71-bf5f-981e40fca1df | 1,272 | 1,512 | Not required | -1 | -1 | turns | the rear tire | 1 | 5ba787ab-4634-451a-b988-c467f4a7fccb | 1,272 | 1,512 | train |
13eaa90a-9292-4eec-8f3a-4b4605a01630 | 8,616 | 8,856 | Not required | -1 | -1 | squeezes | water | 1 | c019e4c6-45f5-4e01-9e4f-6ec4712850ce | 8,615 | 8,855 | train |
9d58bb9b-0905-4499-a40b-12d659739030 | 1,754 | 1,994 | Not required | -1 | -1 | cuts | the plastic | 0 | 130e4f24-c55c-4d09-a1fc-7d9198ae1030 | 1,754 | 1,994 | train |
6f082d5d-5f31-4358-b2dc-16d320312ab3 | 2,369 | 2,609 | Not required | -1 | -1 | holds | the shredded dough | 1 | f938bcd9-bf30-4dfb-9a99-d6b9ee53c046 | 2,368 | 2,608 | train |
Mistake Attribution: Fine-Grained Mistake Understanding in Egocentric Videos
CVPR 2026
Yayuan Li1, Aadit Jain1, Filippos Bellos1, Jason J. Corso1,2
1University of Michigan, 2Voxel51
[Paper] [Code] [Project Page]
MATT-Bench Overview
MATT-Bench provides large-scale benchmarks for Mistake Attribution (MATT) — a task that goes beyond binary mistake detection to attribute what semantic role was violated, when the mistake became irreversible (Point-of-No-Return), and where the mistake occurred in the frame.
The benchmarks are constructed by MisEngine, a data engine that automatically creates mistake samples with attribution-rich annotations from existing egocentric action datasets:
| Dataset | Samples | Instruction Texts | Semantic | Temporal | Spatial |
|---|---|---|---|---|---|
| Ego4D-M | 220,800 | 19,467 | ✓ | ✓ | ✓ |
| EPIC-KITCHENS-M | 299,715 | 12,283 | ✓ | — | — |
These are at least two orders of magnitude larger than any existing mistake dataset. Instruction-text counts = unique (predicate V, argument ARG1) pairs.
A third source, HoloAssist-M, is released alongside as an additional benchmark — see Extended: HoloAssist-M below.
Repository Layout
MATT-Bench/
├── ego4d/
│ ├── train.xlsx, valid.xlsx, test.xlsx ← primary annotation files (consumed by the MATT codebase)
│ ├── parquet.xlsx ← MisEngine reproduction data (Ego4D narrations with SRL)
│ └── parquet/ ← Parquet mirror for the HF dataset viewer
├── epickitchens/
│ ├── train.xlsx, validation.xlsx
│ └── parquet/
└── holoassist/
├── train.xlsx, validation.xlsx
└── parquet/
.xlsx is the canonical download format (the MATT codebase reads Excel directly). The parquet/ mirror powers the HF dataset viewer and datasets.load_dataset(...) loaders — both views contain the same rows.
Downloading MATT-Bench
MATT-Bench has two parts that you obtain separately:
- Annotations — semantic attribution annotations are hosted here, download via
hforgit clone. Temporal and spatial attribution annotations are inherited from the original dataset. - Video media — not hosted here. Download from each source dataset using the instructions below. Original videos retain their upstream licenses.
Annotations (this repo)
# Everything
hf download mistakeattribution/MATT-Bench --repo-type dataset --local-dir MATT-Bench
# Just one source dataset's xlsx files
hf download mistakeattribution/MATT-Bench --repo-type dataset \
--include "ego4d/*.xlsx" --local-dir MATT-Bench
Or via the datasets library (reads the parquet mirror):
from datasets import load_dataset
ego4d_m = load_dataset("mistakeattribution/MATT-Bench", "ego4d")
epic_m = load_dataset("mistakeattribution/MATT-Bench", "epickitchens")
holo_m = load_dataset("mistakeattribution/MATT-Bench", "holoassist")
Video media
Ego4D
Follow https://ego4d-data.org/docs/CLI/ to download. The video_uid and clip1_uid fields in our annotations correspond to Ego4D's native video and clip UIDs.
MATT-Bench uses the FHO (Forecasting Hands and Objects) benchmark clips from Ego4D. Example downloading script:
ego4d --output_directory="~/ego4d_data" --datasets clips --benchmarks FHO
EPIC-KITCHENS-100
Follow https://epic-kitchens.github.io/ to download. MATT-Bench's video_id matches EPIC's participant-video identifier (e.g. P22_16); start_frame / end_frame index the RGB frame sequence.
Example download script:
git clone https://github.com/epic-kitchens/epic-kitchens-download-scripts
cd epic-kitchens-download-scripts
python epic_downloader.py --rgb-frames # or --videos
HoloAssist
Although not reported in the paper, we also support the HoloAssist dataset.
Download the following from the HoloAssist project page:
| Resource | Link | Size |
|---|---|---|
| Videos (pitch-shifted) | video_pitch_shifted.tar | 184.20 GB |
| Labels | data-annotation-trainval-v1_1.json | 111 MB |
| Dataset splits | data-splits-v1_2.zip | — |
MATT-Bench's video_id matches HoloAssist's video identifier (e.g. R076-21July-DSLR).
Data Schema
ego4d/{train,valid,test}.xlsx — 13 columns
| Column | Description |
|---|---|
video_uid |
Ego4D video UID (full video) |
start_frame, end_frame |
Frame bounds of the attempt clip |
clip1_uid, clip1_start_frame, clip1_end_frame |
Primary Ego4D clip |
clip2_uid, clip2_start_frame, clip2_end_frame |
Some actions are distributed across two clips (Not required / -1 when absent) |
V, ARG1 |
Predicate and argument from the instruction (e.g. pick up, apple) |
label |
Mistake label. 0: Correct; 1: Mistaken Predicate; 2: Mistaken Object; 3: Mistaken Both |
split |
dataset split identifier |
ego4d/parquet.xlsx — 29 columns (MisEngine reproduction data)
Ego4D narration-level records with semantic-role labels (ARG0, V, ARG1), frame/time bounds (start_frame/end_frame/start_sec/end_sec), clip-relative bounds, and noun/verb embedding vectors. Used to reproduce the MisEngine step that produces the split files above.
epickitchens/{train,validation}.xlsx and holoassist/{train,validation}.xlsx — 8 columns
| Column | Description |
|---|---|
video_id |
Source-dataset video identifier |
start_frame, end_frame |
Frame bounds of the attempt clip |
V, ARG1 |
Predicate and argument of the instruction text |
label |
Mistake label |
actual_V, actual_ARG1 |
Predicate/argument of the action performed in the video |
Extended: HoloAssist-M
HoloAssist-M is an additional MATT benchmark released alongside MATT-Bench. It is not part of the main two-dataset evaluation reported in the CVPR 2026 paper; it uses the same MisEngine pipeline applied to the HoloAssist dataset.
| Dataset | Samples | Instruction Texts | Semantic | Temporal | Spatial |
|---|---|---|---|---|---|
| HoloAssist-M | 562,209 | 1,786 | ✓ | — | — |
Schema matches EPIC-KITCHENS-M (semantic attribution only — HoloAssist does not provide native PNR frame number andb bbox annotations).
Citation
@inproceedings{li2026mistakeattribution,
title = {Mistake Attribution: Fine-Grained Mistake Understanding in Egocentric Videos},
author = {Li, Yayuan and Jain, Aadit and Bellos, Filippos and Corso, Jason J.},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2026},
}
Please also cite the source datasets:
@inproceedings{grauman2022ego4d,
title = {Ego4D: Around the World in 3,000 Hours of Egocentric Video},
author = {Grauman, Kristen and others},
booktitle = {CVPR},
year = {2022}
}
@article{Damen2022RESCALING,
title = {Rescaling Egocentric Vision: Collection, Pipeline and Challenges for EPIC-KITCHENS-100},
author = {Damen, Dima and others},
journal = {IJCV},
year = {2022}
}
@inproceedings{wang2023holoassist,
title = {HoloAssist: an Egocentric Human Interaction Dataset for Interactive AI Assistants in the Real World},
author = {Wang, Xin and others},
booktitle = {ICCV},
year = {2023}
}
- Downloads last month
- 43