Shows ValueError: WAV header is invalid when implementing voice assistant using Rasa and Mozilla tools

I’m following this blog : Build an AI voice assistant with Rasa Open Source and Mozilla tools | Rasa | The Rasa Blog | Rasa for building a voice assistant and after I run this in the web and give a voice input I get this error:

due to the line in socketio_connector.py: fs, audio = wav.read(“output_{0}.wav”.format(sid)) Is there a way to fix this?

Hi @yachana, how do you capture the voice in the web interface?. Be sure that captured audio is in WAV format, I found that Mozilla getUserMedia only works for OGG formats and you should convert formats if so (e.g. ffmpeg system call or some librosa or pyAudio module).

Note: rasaHQ rasa-voice-interface is quiet oudated and repo web interface javascripts don’t work at once.