Datasets:

Linzhan
/

Objaverse-XL-Rigged-Animated

Derives both CSV indexes from the assets. Pure Python 3, no dependencies: GLB metadata is read from each file's glTF JSON chunk and the binary payload is never touched, so probing the whole 23 GB collection takes about ten seconds.

script	purpose
`glb_probe.py`	Library + CLI. Parses one `.glb` and returns rig, animation and geometry stats.
`build_dataset.py`	Walks every asset once and writes `../metadata.csv` and `../animations.csv`. `--link` also hardlinks a source directory into `../glb/`.

python3 build_dataset.py                      # rebuild both tables in place
python3 build_dataset.py --src /path/raw --link   # ingest a fresh download
python3 glb_probe.py ../glb/<file>.glb        # inspect a single asset

Notes on the probe

Kinematic-tree roots are counted as joints whose parent is not itself a joint, so num_skeleton_roots > 1 marks a disconnected or multi-tree skeleton.
Clip duration comes from the max of each animation sampler's input accessor, which glTF stores in the JSON chunk.
animated_joints intersects a clip's target nodes with the skin's joint set, separating clips that drive a rig from clips that only move object-level nodes.