SCEdit: Efficient and Controllable Image Diffusion Generation via Skip Connection Editing

(CVPR 2024 Highlight)

Zeyinzi Jiang Chaojie Mao Yulin Pan Zhen Han Jingfeng Zhang

Alibaba Group

^* Code and models have been made public on SCEPTER.

Framework

Text-to-image generation is fine-tuned by SC-Tuner, while controllable image synthesis is facilitated by CSC-Tuner and Cascade Dense Convolution. — Text-to-image generation is fine-tuned by SC-Tuner, while controllable image synthesis is facilitated by CSC-Tuner and Cascade Dense Convolution.

Qualitative Results

Citation

@article{jiang2023scedit,
    title = {SCEdit: Efficient and Controllable Image Diffusion Generation via Skip Connection Editing},
    author = {Jiang, Zeyinzi and Mao, Chaojie and Pan, Yulin and Han, Zhen and Zhang, Jingfeng},
    year = {2023},
    journal = {arXiv preprint arXiv:2312.11392}  
}