Human-Voice-Segmentation

This is an algorithm to identify human voice and do segmentation automatically.

The result will be compared to the manual segmentation data, then a accuracy report will be generated based on match rate, insertion rate and omission rate.

1. Download the repository in zip or clone it to your local machine.

2. Add the project into your Matlab directory path.

3. Add you own voice into the folder, and segmentation reference point.

4. Run script3.m then you will get report in the same directory, you might wanna modify the threshold, time tolerance and window size base on your need.