3D-VITAL

Checkpoints of 3D-VITAL: Visibility-aware Identity Training from Artificial Liftings of 3D Reconstruction model for Cross-View Object Re-Identification (NeurIPS 2026): arXiv:2610.16692, poster. Code, training and evaluation are at https://github.com/TurtleSmoke/3D-VITAL.

Checkpoints

Each checkpoint is trained on one camera type of AG-ReID, given by its name.

Checkpoint	Query → Gallery	mAP	Rank-1
agreid_aerial.pth	A → G	52.09	66.20
	G → A	53.30	62.68
agreid_ground.pth	A → G	55.99	67.13
	G → A	57.93	68.61

A = aerial, G = ground. The checkpoints are stored in fp16, so they can differ slightly from the paper.

Usage

git clone https://github.com/TurtleSmoke/3D-VITAL && cd 3D-VITAL && uv sync && source .venv/bin/activate
hf download TurtleSmoke/3D-VITAL --include "*.pth" --local-dir checkpoints
bash scripts/eval.sh agreid_aerial $DATA checkpoints/agreid_aerial.pth

License

The weights are released for non-commercial research under CC BY-NC 4.0. They were trained on AG-ReID and on synthetic views generated with Hunyuan3D-2 and StyleID, whose respective terms also apply. The code is under the CeCILL-B license.

Citation

@inproceedings{grolleau2026vital,
    title = {3D-VITAL: Visibility-aware Identity Training from Artificial Liftings of 3D Reconstruction model for Cross-View Object Re-Identification},
    author = {Grolleau, William and Sabourin, Astrid and Lapouge, Guillaume and Achard, Catherine},
    booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
    year = {2026}
}