Learning stackable and skippable LEGO bricks for efficient, reconfigurable, and variable-resolution diffusion modeling.

Huangjie Zheng^1,2, Zhendong Wang¹, Jianbo Yuan², Guanghan Ning²,
Pengcheng He³, Quanzeng You², Hongxia Yang², Mingyuan Zhou³

¹The University of Texas at Austin, ²ByteDance Inc., ³Microsoft Azure AI

Paper arXiv Code

Hugging Face

BibTeX


      @inproceedings{
        zheng2024learning,
        title={Learning Stackable and Skippable {LEGO} Bricks for Efficient, Reconfigurable, and Variable-Resolution Diffusion Modeling},
        author={Huangjie Zheng and Zhendong Wang and Jianbo Yuan and Guanghan Ning and Pengcheng He and Quanzeng You and Hongxia Yang and Mingyuan Zhou},
        booktitle={The Twelfth International Conference on Learning Representations},
        year={2024},
        url={https://openreview.net/forum?id=qmXedvwrT1}
        }