Valkyra-Labs/charon-audio

Rust music source separation pipeline for ONNX models. Charon runs the HTDemucs 4-stem model (drums, bass, other, vocals) through ONNX Runtime, with decoding, STFT/iSTFT, segmentation, overlap-add and output writing done in Rust. No Python at run time.

语言：Rust · GitHub 星标：18 · Fork：5

在 GitHub 上查看原仓库