demucs
Code for the paper Hybrid Spectrogram and Waveform Source Separation
AI summary
Audio source separator
A deep learning model that separates multiple audio sources from mixed music tracks
- stars
- 8.5K
- forks
- 1.1K
- watching
- 155
Similar projects
Found by comparing what the projects do, not just their names.
Demixing library
A repository containing pre-trained models and code for music demixing using the Hybrid Demucs model
Audio codec
A deep learning-based audio codec that supports high-fidelity neural audio compression.
Music Generation Model
Implementation of Google's MusicLM model for music generation using attention networks and text-conditioning.
Music source separation
A framework for training deep neural networks to separate music sources from audio files.
Transformer platform
A codebase for working with Open Pre-trained Transformers, enabling deployment and fine-tuning of transformer models on various platforms.
Singing voice separator
This project provides a PyTorch implementation of a singing voice separation algorithm using recurrent inference and skip-filtering connections.
Object Detector
A platform for object detection and segmentation tasks using machine learning algorithms
Data generator
This project provides tools and models to generate realistic data for camera systems with defocus blur, aiming to improve image deblurring techniques.
Music generator
Generates music based on user input prompts using the Mubert API
Audio feature extractor
A deep learning tool library for extracting features from audio signals.
Sound classification
Classify music genre from a 10-second sound stream using a neural network.
Avatar generation
Generating photorealistic avatars from audio
Image deblurring algorithm
Deblurring technique using deep learning and Fourier transformation to remove image blur
Multimodal framework
A modular framework for building vision and language multimodal research projects using PyTorch.
Language Model
This repository provides an end-to-end language model capable of generating coherent text based on both spoken and written inputs.
