UIKA: Fast Universal Head Avatar from Pose-Free Images

[Paper] | [Homepage]

We present UIKA, a feed-forward animatable Gaussian head model from an arbitrary number of pose-free inputs, including a single image, multi-view captures, and smartphone-captured videos.

Release Plan

We are in our company's internal review and compliance process for open-sourcing this project. We plan to release the following components before the end of April 2026:

• Testing and training code
• Model checkpoints
• Data generation pipeline

We will update this repository and mark each item as released as soon as it is available.

Acknowledgement

This work is built on many amazing research works and open-source projects:

• LAM
• OpenLRM
• VHAP
• Pixel3DMM
• VGGT

Thanks for their excellent works and great contribution.

Citation

If you find this project useful, please cite:

@inproceedings{wu2026uika,
    title     = {UIKA: Fast Universal Head Avatar from Pose-Free Images},
    author    = {Zijian Wu and Boyao Zhou and Liangxiao Hu and Hongyu Liu and Yuan Sun and Xuan Wang and Xun Cao and Yujun Shen and Hao Zhu},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
    year      = {2026}
}