Defense Date

2021

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Business

First Advisor

Dr. Victoria Yoon

Second Advisor

Dr. Kweku-Muata Osei-Bryson

Third Advisor

Dr. Paul Brooks

Fourth Advisor

Dr. Yeongin Kim

Abstract

Biological languages of the body through which human emotion can be detected abound including heart rate, facial expressions, movement of the eyelids and dilation of the eyes, body postures, skin conductance, and even the speech we make. Speech emotion recognition research started some three decades ago, and the popular Interspeech Emotion Challenge has helped to propagate this research area. However, most speech recognition research is focused on adults and there is very little research on child speech. This dissertation is a description of the development and evaluation of a child speech emotion recognition framework. The higher-level components of the framework are designed to sort and separate speech based on the speaker’s age, ensuring that focus is only on speeches made by children. The framework uses Baddeley’s Theory of Working Memory to model a Working Memory Recurrent Network that can process and recognize emotions from speech. Baddeley’s Theory of Working Memory offers one of the best explanations on how the human brain holds and manipulates temporary information which is very crucial in the development of neural networks that learns effectively. Experiments were designed and performed to provide answers to the research questions, evaluate the proposed framework, and benchmark the performance of the framework with other methods. Satisfactory results were obtained from the experiments and in many cases, our framework was able to outperform other popular approaches. This study has implications for various applications of child speech emotion recognition such as child abuse detection and child learning robots.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

7-24-2021

Share

COinS