wsdt / python_voice_chatbot Goto Github PK

contributing.md to encourage other people

Voice response currently terrible (bad pronounciation, etc.)

By running the current Dockerfile you will get following error:

-> Error opening audio device (null) for capture: Connection refused
In main.py line 14 (main()): liveSpeech() [VoiceMgr.py]

From perspective the microphone or/and loudspeakers of the host system are not mapped into the container.

dockerignore for making the image smaller

Add threading when it comes e.g. to the assistant response (so the we don't stop listen to a users' voice when we wait for the answer)

Better voice-to-text recognition needed.

Maybe add the possibility to use other (cloud-based) apis for those people who do not care much for their privacy or similar.

Maybe migrate CONFIDENTIAL.py (at least for docker images) to docker secrets.