fpaupier / rnnoise

A dockerized version of Recurrent neural network for audio noise reduction

François Paupier 💡

Senior Software Engineer with a focus on Machine Learning engineering and an advocate for results-driven solutions. Leveraging a proven track record of designing and deploying ML workflows tailored to business needs.

🛠 Specializations:

• Machine Learning Workflow Design
• Backend Development with Python (fastAPI, gRPC)
• Frontend Development using React
• Cloud Solutions: AWS, GCP

📚 Education: Centrale Nantes - Master of Science: Signal Processing & Statistics.

🗂 Professional Journey:

• Freelance ML Engineer
• ML engineer at Schneider Electric, Boston, MA
• Data Engineer at Kpler, Paris

🔍 Learning: Kubernetes

💼 Contact: Drop me a message on Twitter or email francois [dot] paupier [at] orange [dot] frfor business inquiries.

Featured Content 📝

• 📕 Blog Posts:

  • Apache NiFi - Surf on Your Dataflow
  • Multiprocessing with Python & gRPC
  • Telegram Bot in Go

• 🎙 Podcast - Post Mortem Podcast: A dive into the intricacies of system outages, exploring causes and solutions. Available on Apple Podcasts, Spotify, Deezer, and Google Podcasts.

Tech Stack 🛠️

My favorite tool is the one that's right for the job at hand. Still, I've developed a particular affinity for some.

📊 Github Analytics:

Connect with me 🔗

rnnoise's Issues

Use alpine instead of ubuntu

to shrink resulting image size

Handle different type of audio inputs

Currently, only wav data, stereo, with a sampling frequency of 48hz and a 16-bit depth are supported.

Extend support to most common audio files, at least wav (all types mono stereo, different sampling rate) and mp3.

See if FFmpeg can not do most of this preprocessing hard work.