ComfyUI-Wan22FMLF (github.com)

The release of ComfyUI-Wan22FMLF introduces a custom node set for the Wan2.2 A14B I2V models, enhancing multi-frame reference conditioning. This update includes significant optimizations, such as injecting the last sampled frame (motion_frames) directly into the latent space for improved frame continuity and using a start image to provide visual guidance. The integration of structural_repulsion_boost allows users to enhance motion effects with a parameter range of 1.0 to 2.0, focusing on preserving color stability in lower noise phases, which is crucial for high-quality video generation. This update holds considerable importance for the AI/ML community as it addresses several performance issues, such as inter-frame discontinuity and reduced file size through server-side storage, enhancing node responsiveness and usability. The revamped conditioning logic and the introduction of flexible reference frame positioning cater to diverse video generation needs, making it easier for developers to create seamless animations with reduced flickering effects. Overall, these enhancements significantly improve the user experience and the versatility of video generation with AI, opening new avenues for creative applications in the field.