Skip to content
ModelListedExternal
Download

Seg Head Vspw

Segmentation head trained on VSPW (mIoU: 58.4). Part of A Frame is Worth One Token: Efficient Generative World Modeling with Delta Tokens (CVPR 2026 Highlight).

Sourcehuggingface.co/Amazon-FAR/seg-head-vspw/tree/a872942158200d808406f305dde0640535f91b45
KindOtherVersionRevision a872942

Model source

Kind
Other
File size
388 KB
Version
vRevision a872942
Updated
Sep 22, 2026
Downloads
0
Source
Hugging Face

Source description

Segmentation Head — VSPW

Segmentation head trained on VSPW (mIoU: 58.4). Part of A Frame is Worth One Token: Efficient Generative World Modeling with Delta Tokens (CVPR 2026 Highlight).

Usage

Requires a frozen DINOv3 ViT-B backbone. See the DeltaTok GitHub repository for training and evaluation code.

Acknowledgements

Citation

@inproceedings{kerssies2026deltatok,
  title     = {A Frame is Worth One Token: Efficient Generative World Modeling with Delta Tokens},
  author    = {Kerssies, Tommie and Berton, Gabriele and He, Ju and Yu, Qihang and Ma, Wufei and de Geus, Daan and Dubbelman, Gijs and Chen, Liang-Chieh},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year      = {2026}
}

Sources

1 source
Hugging FacePrimaryVersion a872942158200d808406f305dde0640535f91b45
huggingface.co/Amazon-FAR/seg-head-vspw/tree/a872942158200d808406f305dde0640535f91b45

Verified Sep 8

Model artifacts

1 artifact

pytorch_model.bin

bin · 388 KB · SHA-256 d9409ab321f9…48db · Hugging Face

Download

Source excerpts

2 excerpts
ReadmeSep 8huggingface.co/Amazon-FAR/seg-head-vspw

--- datasets: - vspw library_name: pytorch license: apache-2.0 pipeline_tag: image-segmentation tags: - deltatok - cvpr2026-highlight --- # Segmentation Head — VSPW Segmentation head trained on VSPW (mIoU: 58.4). Part of [A Frame is Worth One Token: Efficient Generative World Modeling with Delta Tokens](https://huggingface.co/papers/2604.04913) (CVPR 2026 Highlight). ## Usage Requires a frozen [DINOv3](https://github.com/facebookresearch/dinov3) ViT-B backbone. See the [DeltaTok GitHub repository](https://github.com/amazon-far/deltatok) for training and evaluation code. ## Acknowledgements - [DINOv3](https://github.com/facebookresearch/dinov3) - [VSPW](https://www.vspwdataset.com/) ## Citation ```bibtex @inproceedings{kerssies2026deltatok, title = {A Frame is Worth One Token: Efficient Generative World Modeling with Delta Tokens}, author = {Kerssies, Tommie and Berton, Gabriele and He, Ju and Yu, Qihang and Ma, Wufei and de Geus, Daan and Dubbelman, Gijs and Chen, Liang-Chieh}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, year = {2026} } ```

NoteSep 8huggingface.co/Amazon-FAR/seg-head-vspw

Source context: 0 downloads · 1 likes · Pipeline image-segmentation · Library pytorch · Repo Amazon-FAR/seg-head-vspw