AI-based Background Noise Detection and Auto-Mute Feature

[Madhusri1019]

Hello Jitsi team

I'd like to suggest a new project idea for GSoC 2025 that enhances the audio experience in Jitsi Meet by intelligently detecting background noise and helping users manage it.

Title:
AI-based Background Noise Detection and Auto-Mute Feature

** Description:**
The goal of this project is to develop a smart audio module that can identify disruptive background sounds (such as keyboard typing, barking dogs, or traffic) in real time and automatically mute the user or suggest a mute action.

Key features include:

1. Lightweight background noise classifier (client-side or server-side)

2. UI integration for mute suggestions or auto-mute

3.Configurable sensitivity and toggle controls

4.Seamless integration with existing Jitsi Meet audio pipeline

5.This would be especially useful for large meetings, educational settings, and public webinars.

Benefits to Jitsi:
1 . Improved meeting experience and professionalism

2. Reduced distractions in group calls

3. Adds intelligent features to support accessibility and usability

Technologies:
Java, JavaScript, TensorFlow.js or similar ML toolkit, WebRTC, lib-jitsi-meet

** Difficulty:**
Medium

Would really appreciate any thoughts or feedback on this! I’m excited about the potential to contribute to Jitsi through GSoC 2025 and would love to refine this idea further with the community's input.

Thanks!