https://github.com/maxadams0/split-audio

Split audio by sentence for use in AI Voice Models
https://github.com/maxadams0/split-audio

Last synced: about 2 months ago
JSON representation

Split audio by sentence for use in AI Voice Models

• Host: GitHub
• URL: https://github.com/maxadams0/split-audio
• Owner: MaxAdams0
• License: gpl-3.0
• Created: 2023-08-06T01:40:43.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-08-07T04:22:31.000Z (over 1 year ago)
• Last Synced: 2023-08-07T04:42:10.323Z (over 1 year ago)
• Language: Python
• Size: 17.6 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Split-Audio

A quick Python script to detect speach using Whisper, parse it by sentance, and then output each sentance into a new audio file. This is intended for use in AI Model Training, but use it in whatever project you like! Please make sure to abide by GPLv3 License rules.

## Requirements

The only requirements are related to installs on your computer, any required python libraries for the script will be installed automatically using setup.py.

- [Python](https://www.python.org/downloads/release/python-31011/) (only 3.10.11 was tested, others may work)

- [FFmpeg](https://ffmpeg.org/download.html)

***This script is meant for windows only***

You may need to add either of these to your PATH Environmental Variables

## Use

1. Run `setup.py`, and wait until it is done

2. Put your audio file(s) into the input folder

3. Run `start.bat` and wait

## Bugs

- PyTorch cannot correctly find gpu index if run through python file directly, unknown why (hense batch file)

## Future Updates

These are in no specific order and may be completed at different times... or never

- Add ***diarization*** (using [NVIDIA NeMo](https://github.com/NVIDIA/NeMo/tree/main/examples/speaker_tasks/diarization)

- Add error handling

- Add logging

## Changelog