Voicepen project live : https://gdsc-voicepen.netlify.app/

Demo video P1 : Link , P2 : Link

How it works :

• This project utilises huggingface api for speech recognition and summarization - link
• First, it records audio and sends it to the speech recognition API.
• Audio recording is implemented using the React Audio Voice Recorder NPM package: react-audio-voice-recorder
• If there is already an audio file available, it can be selected and sent for speech recognition instead.
• Once the speech is recognized and text is sent back, it is sent to the summarization API to get a summary back.

V2 : Variant of this

• Instead of recording audio and then sending it to speech recognition, this version tries to do live speech recognition.
• This functionality is implemented using an npm package called react-speech-recognition
• After you speak something, clicking the "Summarize" button sends the recognized speech to the summarization API.

Techstack

• React JS - Vite