Datasets:
Size:
1K - 10K
ArXiv:
Tags:
humanoid-locomanipulation
whole-body-control
human-object-interaction
video-to-motion
reinforcement-learning
physics-simulation
License:
update README
Browse files
README.md
CHANGED
|
@@ -14,6 +14,14 @@ tags:
|
|
| 14 |
library_name: GRAIL
|
| 15 |
---
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
# Dataset Overview
|
| 18 |
|
| 19 |
| Tabletop Pickup | Ground Pickup |
|
|
@@ -83,6 +91,7 @@ nvidia/PhysicalAI-Robotics-Locomanipulation-GRAIL/
|
|
| 83 |
β βββ meta/<hoi_category>__<object>__<NNN>.pkl # per-motion metadata (lengths, contact flags, source IDs)
|
| 84 |
β βββ object_usd/<hoi_category>__<object>__<NNN>.usd # OpenUSD object asset
|
| 85 |
β β βββ textures/<basename>/ # per-USD texture subdir, refs rewritten in the USD
|
|
|
|
| 86 |
βββ checkpoint/
|
| 87 |
βββ GEM-SMPL/ # SMPL-X human pose estimation weights (HMR2, ViTPose, VIMO, YOLO, HMR4D)
|
| 88 |
βββ FoundationPose/weights/ # object 6-DoF estimator (refiner + scorer)
|
|
|
|
| 14 |
library_name: GRAIL
|
| 15 |
---
|
| 16 |
|
| 17 |
+
## π’ News
|
| 18 |
+
|
| 19 |
+
- **[2026-07-15]** Released task-general tracking policy checkpoints trained on the released data. Follow the [tracking doc](https://nvlabs.github.io/GRAIL/tracking.html#tracking-released-data) to use them to track our released motion data.
|
| 20 |
+
|
| 21 |
+
> [!WARNING]
|
| 22 |
+
> `data/pickup_table` and `data/pickup_ground` were updated on 2026-07-14. If you downloaded it before then, please re-download it.
|
| 23 |
+
|
| 24 |
+
|
| 25 |
# Dataset Overview
|
| 26 |
|
| 27 |
| Tabletop Pickup | Ground Pickup |
|
|
|
|
| 91 |
β βββ meta/<hoi_category>__<object>__<NNN>.pkl # per-motion metadata (lengths, contact flags, source IDs)
|
| 92 |
β βββ object_usd/<hoi_category>__<object>__<NNN>.usd # OpenUSD object asset
|
| 93 |
β β βββ textures/<basename>/ # per-USD texture subdir, refs rewritten in the USD
|
| 94 |
+
β βββ bps/<hoi_category>__<object>__<NNN>.npy # BPS encoding of object 3D shape (only in pickup table/ground)
|
| 95 |
βββ checkpoint/
|
| 96 |
βββ GEM-SMPL/ # SMPL-X human pose estimation weights (HMR2, ViTPose, VIMO, YOLO, HMR4D)
|
| 97 |
βββ FoundationPose/weights/ # object 6-DoF estimator (refiner + scorer)
|