useAudioDetection

[ttessarolo]

Hi I've rewritten your code using and audioWorklet from Chromium team. I've also created a React hook:

https://github.com/ttessarolo/useAudioDetection

[solyarisoftware]

Ciao Tommaso,
It would be great if you could mention my project as well! I kindly invite you to refer to it.

I’m not deeply involved in this JavaScript work anymore, but I’m curious to understand why you decided to port it to React—there are probably good reasons behind that. At the time, my goal was to deliberately write it in vanilla JavaScript.

Adding an explanatory note and a kind reference to the original project would be much appreciated.

Thank you,
Giorgio

[ttessarolo]

Ciao Giorgio,

all done: I've included a Reference paragraph.

Regarding the motivations, I was exploring various approaches to send live audio to an LLM model. First, I tried evaluating solutions that used the client for audio chunking. Along the way, I radically changed my approach, opting for WebRTC and server-side processing instead. Check this out –> https://github.com/ttessarolo/ai-realtime-api. I’m trying to recreate OpenAI’s real-time API to use with local models.