docs: WorldRover -> WorldRover-preview
Browse files
README.md
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
---
|
| 2 |
license: other
|
| 3 |
license_name: worldrover-research
|
| 4 |
-
pretty_name: WorldRover
|
| 5 |
task_categories:
|
| 6 |
- depth-estimation
|
| 7 |
- robotics
|
|
@@ -18,21 +18,23 @@ size_categories:
|
|
| 18 |
- 100B<n<1T
|
| 19 |
---
|
| 20 |
|
| 21 |
-
# WorldRover
|
| 22 |
|
| 23 |
Long, continuous camera paths through photoreal 3D environments, rendered with **per-frame
|
| 24 |
metric depth, camera pose and action labels** — and, where it matters, rendered **more than
|
| 25 |
once along the same path**, so you can hold geometry and motion fixed while the projection or
|
| 26 |
the lighting changes.
|
| 27 |
|
| 28 |
-
This repository is the **
|
| 29 |
-
|
|
|
|
|
|
|
| 30 |
|
| 31 |
## Parts
|
| 32 |
|
| 33 |
| Repository | What it is | Scenes | Clips | Video | Size |
|
| 34 |
|---|---|---|---|---|---|
|
| 35 |
-
| **this repo** | lite: paired panoramic + first-person, RGB + pose + actions, **no depth** | 4 | 129 per view | 4.1 h per view | 103 GB |
|
| 36 |
| [**WorldRover-6scenes**](https://huggingface.co/datasets/AlayaLab/WorldRover-6scenes) | paired 360° panoramic + first-person, **with lossless depth** | 6 | 600 per view | 18.9 h per view | 7.6 TB |
|
| 37 |
| [**WorldRover-styles**](https://huggingface.co/datasets/AlayaLab/WorldRover-styles) | one trajectory set re-rendered under 6 lighting/style treatments | 4 | 1000 | 48 h | 818 GB |
|
| 38 |
| [med_village](https://huggingface.co/datasets/AlayaLab/WorldRover-med_village) · [paris](https://huggingface.co/datasets/AlayaLab/WorldRover-paris) · [venice](https://huggingface.co/datasets/AlayaLab/WorldRover-venice) · [art_nouveau](https://huggingface.co/datasets/AlayaLab/WorldRover-art_nouveau) | the original release, full depth, one repo per scene | 1 each | 13–47 per view | ~30 min per view | 169–334 GB |
|
|
@@ -42,7 +44,7 @@ release or the style variants — it is the largest, newest and most complete pa
|
|
| 42 |
|
| 43 |
## The two kinds of pairing
|
| 44 |
|
| 45 |
-
**Same path, two projections** (`WorldRover-6scenes`, and this
|
| 46 |
`fp/<clip_id>` are the same camera path: the first-person clip is rendered from the panoramic
|
| 47 |
clip's per-frame trajectory, so the two `camera_trajectory.csv` files agree row for row and only
|
| 48 |
the intrinsics differ (360°/0 mm equirect vs 65.5°/28 mm pinhole). No interpolated alignment is
|
|
|
|
| 1 |
---
|
| 2 |
license: other
|
| 3 |
license_name: worldrover-research
|
| 4 |
+
pretty_name: WorldRover (preview)
|
| 5 |
task_categories:
|
| 6 |
- depth-estimation
|
| 7 |
- robotics
|
|
|
|
| 18 |
- 100B<n<1T
|
| 19 |
---
|
| 20 |
|
| 21 |
+
# WorldRover-preview
|
| 22 |
|
| 23 |
Long, continuous camera paths through photoreal 3D environments, rendered with **per-frame
|
| 24 |
metric depth, camera pose and action labels** — and, where it matters, rendered **more than
|
| 25 |
once along the same path**, so you can hold geometry and motion fixed while the projection or
|
| 26 |
the lighting changes.
|
| 27 |
|
| 28 |
+
This repository is the **original preview release**, as a lite subset: RGB, camera pose,
|
| 29 |
+
actions and metadata for four scenes, **without depth** (depth is 87% of the bytes). The
|
| 30 |
+
rest of the data lives in the repositories below — see the [collection](https://huggingface.co/collections/AlayaLab/worldrover-6ab4ee625ea9252fbae26a66) for all
|
| 31 |
+
parts in one place.
|
| 32 |
|
| 33 |
## Parts
|
| 34 |
|
| 35 |
| Repository | What it is | Scenes | Clips | Video | Size |
|
| 36 |
|---|---|---|---|---|---|
|
| 37 |
+
| **this repo** (`WorldRover-preview`) | lite: paired panoramic + first-person, RGB + pose + actions, **no depth** | 4 | 129 per view | 4.1 h per view | 103 GB |
|
| 38 |
| [**WorldRover-6scenes**](https://huggingface.co/datasets/AlayaLab/WorldRover-6scenes) | paired 360° panoramic + first-person, **with lossless depth** | 6 | 600 per view | 18.9 h per view | 7.6 TB |
|
| 39 |
| [**WorldRover-styles**](https://huggingface.co/datasets/AlayaLab/WorldRover-styles) | one trajectory set re-rendered under 6 lighting/style treatments | 4 | 1000 | 48 h | 818 GB |
|
| 40 |
| [med_village](https://huggingface.co/datasets/AlayaLab/WorldRover-med_village) · [paris](https://huggingface.co/datasets/AlayaLab/WorldRover-paris) · [venice](https://huggingface.co/datasets/AlayaLab/WorldRover-venice) · [art_nouveau](https://huggingface.co/datasets/AlayaLab/WorldRover-art_nouveau) | the original release, full depth, one repo per scene | 1 each | 13–47 per view | ~30 min per view | 169–334 GB |
|
|
|
|
| 44 |
|
| 45 |
## The two kinds of pairing
|
| 46 |
|
| 47 |
+
**Same path, two projections** (`WorldRover-6scenes`, and this preview repo). `pano/<clip_id>` and
|
| 48 |
`fp/<clip_id>` are the same camera path: the first-person clip is rendered from the panoramic
|
| 49 |
clip's per-frame trajectory, so the two `camera_trajectory.csv` files agree row for row and only
|
| 50 |
the intrinsics differ (360°/0 mm equirect vs 65.5°/28 mm pinhole). No interpolated alignment is
|