Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -97,8 +97,9 @@ All repos live under [`MIT-Media-Lab`](https://huggingface.co/MIT-Media-Lab) and
|
|
| 97 |
| **Something-Something V2** | 52,706 | 1,124,722 | text only | 4.63 GB | [`ssv2-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/ssv2-annotations-v1) |
|
| 98 |
| **OakInk2** | 28,264 | 1,371,721 | episodes + text | 1.92 GB | [`oakink2-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/oakink2-annotations-v1) |
|
| 99 |
| **TACO** | 23,757 | 736,136 | episodes + text | 1.34 GB | [`taco-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/taco-annotations-v1) |
|
|
|
|
| 100 |
| **H2O** | 5,696 | 196,941 | episodes + text | 0.40 GB | [`h2o-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/h2o-annotations-v1) |
|
| 101 |
-
| **total** | **
|
| 102 |
|
| 103 |
**episodes** = entries in `episode_frame_index.npz`, i.e. what a training run actually sees.
|
| 104 |
**training samples** = rows of `index_frame_pair`; the row number *is* the sample id.
|
|
@@ -120,6 +121,7 @@ instruction. That is why the published counts are below the totals we cut:
|
|
| 120 |
| Something-Something V2 | 52,706 | 52,706 | 0 |
|
| 121 |
| OakInk2 | 37,692 | 28,264 | 9,427 (25.0%) |
|
| 122 |
| TACO | 26,454 | 23,757 | 2,697 (10.2%) |
|
|
|
|
| 123 |
| H2O | 7,792 | 5,696 | 2,096 (26.9%) |
|
| 124 |
|
| 125 |
`ssv2` and `egoexo4d` are 0 because VITRA-1M already dropped `N/A` upstream — their episodes
|
|
@@ -139,7 +141,6 @@ head-mounted camera at all, so there was nothing to filter down to.
|
|
| 139 |
|
| 140 |
| dataset | episodes cut | where it stands |
|
| 141 |
|---|---:|---|
|
| 142 |
-
| ARCTIC | 14,826 | cut only — no rendering, captions or index yet |
|
| 143 |
| HOI4D | — | source converted by a colleague; not re-cut |
|
| 144 |
| HOT3D | — | source converted by a colleague; not re-cut |
|
| 145 |
<!-- COLLECTION_TABLE_END -->
|
|
|
|
| 97 |
| **Something-Something V2** | 52,706 | 1,124,722 | text only | 4.63 GB | [`ssv2-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/ssv2-annotations-v1) |
|
| 98 |
| **OakInk2** | 28,264 | 1,371,721 | episodes + text | 1.92 GB | [`oakink2-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/oakink2-annotations-v1) |
|
| 99 |
| **TACO** | 23,757 | 736,136 | episodes + text | 1.34 GB | [`taco-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/taco-annotations-v1) |
|
| 100 |
+
| **ARCTIC** | 12,610 | 425,796 | episodes + text | 0.85 GB | [`arctic-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/arctic-annotations-v1) |
|
| 101 |
| **H2O** | 5,696 | 196,941 | episodes + text | 0.40 GB | [`h2o-annotations-v1`](https://huggingface.co/datasets/MIT-Media-Lab/h2o-annotations-v1) |
|
| 102 |
+
| **total** | **521,299** | **15,585,800** | | **44.9 GB** | |
|
| 103 |
|
| 104 |
**episodes** = entries in `episode_frame_index.npz`, i.e. what a training run actually sees.
|
| 105 |
**training samples** = rows of `index_frame_pair`; the row number *is* the sample id.
|
|
|
|
| 121 |
| Something-Something V2 | 52,706 | 52,706 | 0 |
|
| 122 |
| OakInk2 | 37,692 | 28,264 | 9,427 (25.0%) |
|
| 123 |
| TACO | 26,454 | 23,757 | 2,697 (10.2%) |
|
| 124 |
+
| ARCTIC | 14,821 | 12,610 | 2,211 (14.9%) |
|
| 125 |
| H2O | 7,792 | 5,696 | 2,096 (26.9%) |
|
| 126 |
|
| 127 |
`ssv2` and `egoexo4d` are 0 because VITRA-1M already dropped `N/A` upstream — their episodes
|
|
|
|
| 141 |
|
| 142 |
| dataset | episodes cut | where it stands |
|
| 143 |
|---|---:|---|
|
|
|
|
| 144 |
| HOI4D | — | source converted by a colleague; not re-cut |
|
| 145 |
| HOT3D | — | source converted by a colleague; not re-cut |
|
| 146 |
<!-- COLLECTION_TABLE_END -->
|