Realtime Whisper Flet: Speech-to-Text GUI with Live Transcription 🎙️

[alouiadel]

Hey everyone! I've just open-sourced a project that's been sitting on my hard drive since university days - a modern GUI for Faster Whisper speech recognition using the Flet framework.

ezgif-689e5df45b8aa5.mp4

What it does:

• 🔊 Transcribe audio files with a clean, intuitive UI
• 🎤 Record audio directly and transcribe it on the spot
• 🎧 Live transcription with real-time speech recognition
• 🌍 Support for both English-only and multilingual models
• 🗣️ Translate any language to English text
• 📏 Multiple model sizes with VRAM/speed indicators
• 💻 GPU (CUDA) or CPU processing options
• 🌓 Dark/light theme toggle

The app provides an easy way to use the powerful Whisper speech recognition model without dealing with command-line tools or writing code. It's designed to be accessible while still offering flexibility for power users.

I built this initially for my own use, but figured others might find it helpful too. Would love to hear if you find it useful - any feedback, stars, or contributions are very welcome!

Anyone else working on Whisper-related projects? What use cases are you exploring?

Link to repo : https://github.com/alouiadel/whisper-realtime-echo