StoryDiffusion: Consistent Self-Attention for Long-Range Image and Video Generation
Github上的人脸修复、人脸清晰开源项目