Audio Segregation System
- Category: Training Programme
- Client: Hexaware Technologies
- Project date: 15 November, 2018
Project Details
Developed an audio segregation system that aimed to segregate two or more mixed audio sources into separate audio sources using an unsupervised machine learning algorithm called Cocktail Party Algorithm. This algorithm can be used to collect important information from clustered audio sources. This algorithm was inspired by the research of Dr. Andrew Ng on Singular Value Decomposition (SVD) theory.