Implementation of Band Split Roformer, SOTA Attention network for music source separation out of ByteDance AI Labs

KUIELAB-MDX-Net got the 2nd place on the Leaderboard A and the 3rd place on the Leaderboard B in the MDX-Challenge ISMIR 2021

Singing Voice Separation via Recurrent Inference and Skip-Filtering Connections - PyTorch Implementation. Demo:

The code for the MaD TwinNet. Demo page:

C++17 port of Demucs v3 (hybrid) and v4 (hybrid transformer) models with ggml and Eigen3

Pytorch implementation of subband decomposition

An official implementation of the ICASSP 2024 paper: Dual-Path TFC-TDF UNet for Music Source Separation

Sound Demixing Challenge 2023

Control mechanisms to the U-Net architecture for doing multiple source separation instruments

Banquet: A Stem-Agnostic Single-Decoder System for Music Source Separation Beyond Four Stems

Source Separation training codebase for the Sound Demixing Challenge 2023.

Gradio WebUI for whisper, faster-whisper, whisper-timestamped. Supports YouTube Downloader, Vocal Remover and Transcription.

logWMSE, an audio quality metric & loss function with support for digital silence target. Useful for training and evaluating audio source separation systems.

Create and explore isolated tracks from music files

Support material and source code for the system described in : "New Sonorities for Jazz Recordings: Separation and Mixing using Deep Neural Networks".

Support material and source code for the model described in : "A Recurrent Encoder-Decoder Approach With Skip-Filtering Connections For Monaural Singing Voice Separation"

Pytorch implementation of MDensenet and sparse NMF. Made for my undergraduate thesis "Music Source Separation with Supervised Learning Methods".

Audio Signal Processing Python Tools

Carnatic singing voice separation trained with in-domain data with leakage

eCMU: An Efficient Phase-aware Framework for Music Source Separation with Conformer (IEEE RIVF23)