BYU

Abstract by Kolby Nottingham

Personal Infomation


Presenter's Name

Kolby Nottingham

Co-Presenters

None

Degree Level

Undergraduate

Co-Authors

None

Abstract Infomation


Department

Physics and Astronomy

Faculty Advisor

Mark Transtrum

Title

Acoustics Crowd Behavior

Abstract

Advancements in machine learning that evaluate textual and visual data have been a major research focus during the past decade. Attention has been given to processing audio data to analyze music and human speech, but little attention has been given to analyzing crowd noise. Teaching a machine to analyze crowd noise and recognize changes in crowd mood can help crowd control and event analysis. Our team hopes to use machine learning techniques to correctly identify changes in crowd / audience mood during sports games. Our research makes steps in new aspects of machine learning and event audio analysis and, with continued research, will have applications in crowd control and analysis. Here, I analyze the method by which we will apply machine learning to classify crowd audio data, specifically what methods we use to identify classification labels for training data.