calpoly-csai / csai_voice_assistant Goto Github PK

Fix issues with target dir in the audio recording script and MFCC storage in feature extraction scripts.

The user will input a directory or audio file and the program will output a padded audio file to the correct length. The user can also augment data by changing pitch, etc.

Traceback (most recent call last):
  File "Audio_Data_Recording.py", line 14, in <module>
    import winsound
ModuleNotFoundError: No module named 'winsound'

Seems like winsound is specific to the Windows operating system. Could we use a platform agnostic audio module or use an alternate package for MacOS/Linux?

This is gonna be a tough one.

We've got a lot of unit tests to write...

We need a requirements.txt for handling all needed Python libraries

Is there a reason to not ignore all .json files?

Used to read in an audio file and transcribe it to text format.

At some point, once we have a decent model trained, include it as a "default" model in the repository. (add the model file to the repo and add an exception in .gitignore)