Multi Speaker Segmentation

Human beings use speech as a fundamental method of communication, therefore it consists of bio-metric values which can be used for the identification and classification of humans. The project is useful for security organizations, identifying voices in: meetings, conferences and telephone conver

Project Title

Multi Speaker Segmentation

Project Area of Specialization

Artificial Intelligence

Project Summary

Human beings use speech as a fundamental method of communication, therefore it consists of bio-metric values which can be used for the identification and classification of humans. The project is useful for security organizations, identifying voices in: meetings, conferences and telephone conversations. This project provides enhanced and precise method to perform this task using an artificial intelligence algorithm.

Project Objectives

Speaker Segmentation is the task to estimate “who spoke when” from an audio recording. A speaker segmentation system will discover how many people are involved in the meeting. It is a key technology for various applications using Automatic Speech Recognition (ASR) in multi-talker scenarios such as telephone conversations, meetings, conferences and lectures.

Project Implementation Method

The project would be divided into three phases listed below.
• Database Development for Speakers.
• Speaker Diarization Algorithm Development.
• Algorithm Optimization and Improvement for handling bigger datasets.

Benefits of the Project

Multi Speaker Segmentation or what this project is aimed to imitate is the recognition and segmentation of a targeted speaker or speakers using the speech’s bio-metric values. The project has relevance in the field of intelligence gathering, home security systems and even plays a vital in furthering the base of IoT in our lives.

Technical Details of Final Deliverable

The Expected deliverables of the project are as follows
• To Detect the Number of Speakers i.e. to know how many speakers are present in a meeting.
• Segregate Audio Segments corresponding to each Speaker, classify the different number of speakers and their speech accordingly.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

IT

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Quality Education, Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com