#148 · Primary category: Video & Animation

Anime4KCPP

anime anime4k anime4kcpp avisynth avisynthplus-plugin cnn computer-graphics cpp directshow-filter gpu-acceleration machine-learning upscaling vapoursynth vapoursynth-plugin video-processing

A high performance anime upscaler

Project last updated:07/04/26

GitHub Stars

2.0K

Forks

144

Contributors

7

License

GPL-3.0

Why we included this project

Anime4KCPP is a CNN-based upscaler built specifically for anime footage, and it is one of the easier ones to drop into an existing pipeline. It comes as a command-line tool with VapourSynth and AviSynth plugins plus a Windows DirectShow filter, so it fits the filter chains video editors and encoder enthusiasts already use rather than demanding a new workflow. The v3 rewrite retrains the ACNet and ARNet models and adds several variants, and it runs on CUDA, OpenCL, and plain CPU SIMD paths, which is handy when you want GPU acceleration on one machine but portability on another. A browser-based WebAssembly playground lets you preview results before committing to a full encode. Teams maintaining large anime libraries or building automated upscaling scripts will find the Python binding and FFmpeg integration especially convenient.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category