Zijian Zhou^1,2, Shikun Liu¹, Haozhe Liu¹, Haonan Qiu¹, Zhaochong An¹, Weiming Ren¹, Zhiheng Liu¹, Xiaoke Huang¹, Kam Woh Ng¹, Tian Xie¹, Xiao Han¹, Yuren Cong¹, Hang Li¹, Chuyan Zhu¹, Aditya Patel¹, Tao Xiang¹, Sen He¹

¹ Meta AI     ² King's College London

The training and inference code will be released once it has been organized. Please stay tuned.

Saber is a scalable zero-shot framework for reference-to-video (R2V) generation. By introducing a masked training strategy, Saber bypasses the bottleneck of explicit reference image-video-text triplet datasets, training exclusively on video-text pairs to achieve zero-shot generation capabilities without explicit R2V data.

If you find Saber useful for your research, please cite our paper:

@article{zhou2025scaling,
    title={Scaling Zero-Shot Reference-to-Video Generation},
    author={Zhou, Zijian and Liu, Shikun and Liu, Haozhe and Qiu, Haonan and An, Zhaochong and Ren, Weiming and Liu, Zhiheng and Huang, Xiaoke and Ng, Kam Woh and Xie, Tian and Han, Xiao and Cong, Yuren and Li, Hang and Zhu, Chuyan and Patel, Aditya and Xiang, Tao and He, Sen},
    journal={arXiv preprint arXiv:2512.06905},
    year={2025}
}