mpariente
Audio researcher
Funding Links: https://github.com/sponsors/mpariente
- Name: Pariente Manuel
- Location: Sophia Antipolis
- Company: Pulse Audition
- Kind: user
- Followers: 431
- Following: 419
- Total stars: 559
- Repositories count: 55
- Created at: 2022-11-12T16:10:41.915Z
- Updated at: 2025-03-27T22:33:54.831Z
- Last synced at: 2025-03-27T22:33:54.831Z
GitHub Sponsors Profile
Hi 👋, I'm Manu
I write open-source software all the time to solve issues I face as a researcher/developer. Some of my projects are quite used and, I hope, help researchers and developers.
My main projects:
Asteroid - asteroid is the PyTorch-based audio source separation toolkit for researchers. It is starting to be widely adopted in the speech community. It provides free-to-use pretrained models with SOTA performance.
pystoi - pystoi is a Python implementation of the Short Term Objective Intelligibility (STOI) measure originally implemented in Matlab. Still widely used, STOI was only implemented in Matlab. This Python implementation is used in speech_metrics and pb_bss.
pytorch_stoi - torch_stoi is a PyTorch implementation of the STOI measure, suitable to train speech related deep neural networks.
... and contributions to PyTorch, PyTorchLighning, pytorch-optimizer, pytorch-sparse, pytorch-scatter, torch-mir-eval, SoundCard, Ranger and virtually all the repos that I use.
Why Sponsor Me?
Keep it free: if I'm sponsored for the work I do, I won't need to make code closed source or add premium (aka "paid") features in the future.
Keep free support: no one likes answering issues but whether you're a newbie or senior developer at a massive company, sometimes you need to ask a question and get a prompt, intelligent answer.
Support the principle: open-source software has changed the world for the better, contributing to the open-source community financially will help that positive change continue.
- Current Sponsors: 0
- Past Sponsors: 0
- Total Sponsors: 0
- Minimum Sponsorship: $5.00
Featured Works
mpariente/pystoi
Python implementation of the Short Term Objective Intelligibility measure
Language: MATLAB - Stars: 337asteroid-team/torch-audiomentations
Fast audio data augmentation in PyTorch. Inspired by audiomentations. Useful for deep learning.
Language: Python - Stars: 1021mpariente/pytorch_stoi
STOI loss function in PyTorch
Language: Python - Stars: 91Active Sponsors
Past Sponsors
Sponsor Breakdown
- User: 1