Today's featured AI news highlights, happy reading 👇

🎨 ComfyUI is launching its first official client, supporting cross-platform functionality and integrating multiple useful features. Currently requires waitlist application.

🎭 GAGAvatar releases open-source real-time facial expression transfer technology, requiring only one photo and a dynamic expression video to achieve real-time facial expression transfer.

✂️ LosslessCut, an open-source lossless video editing tool, supports quick compression and video clip extraction without re-encoding.

🎵 AI tool UVR5-UI provides simple online vocal and instrumental separation functionality, supporting multiple AI models and batch processing.

AI Art

1. ComfyUI is finally releasing its official client!

Supports Windows, macOS, and Linux systems, featuring code signing, automatic updates, and lightweight performance.

Detailed introduction: https://blog.comfy.org/comfyui-v1-release/

Official website: https://www.comfy.org/

Additionally, the client includes ComfyUI manager, custom nodes, automatic resource import, and integrated log viewing functions.

However, currently you need to apply to join the waitlist to experience it.

Cutting-edge Technology

1. A real-time facial expression transfer technology: GAGAvatar.

By inputting a dynamic facial expression video + a static portrait image, it can make the static portrait perform actions consistent with the dynamic expressions.

Detailed introduction: https://xg-chu.site/project_gagavatar/

GitHub: https://github.com/xg-chu/GAGAvatar

From the demonstrations provided, the results look impressive. The code is now open source, interested users can deploy it and try it out.

Open Source Projects

1. An efficient lossless audio and video editing tool: LosslessCut.

Can complete lossless compression of audio and video within seconds, particularly suitable for high-definition video files from cameras, GoPro, and drones.

GitHub: https://github.com/mifi/lossless-cut

Moreover, it can quickly extract highlights from videos and supports adding music or subtitle tracks without re-encoding.

2. A powerful AI tool for vocal and instrumental separation: UVR5-UI.

Provides a simple and user-friendly online interface, supports selection of different AI models, output formats and quality, and supports batch processing.

GitHub: https://github.com/Eddycrack864/UVR5-UI

Currently available on Colab or HuggingFace, those who need it can give it a try.