Human Speech Component Analysis

Overview

Objectives/Goals This project#s purpose is to explore the components of speech perception, to answer the question #What do humans actually hear when they listen to speech.# This project was done to add more to the speech knowledge base and research methods for speech recognition. Methods/Materials A speech component analysis experiment was conducted. In this experiment, voice samples were filtered so pitch, power, and relative frequency compositions are separated into different samples. The experiment consisted of a multiple choice accuracy of perception test. Results The statistical analysis shows that the fundamental pitch and the relative power of the sound failed as a medium of communication, but the relative pitch compositions allowed subjects to discern almost half of the phrases. Conclusions/Discussion The data supports the original hypothesis. Relative compositions contain much more discernable audio information in relation to speech. The information collected in this project may be used for further speech research. Future algorithmic speech recognition research based on relative harmonics should be conducted.

Competition history

• CSEF 2008 Behavioral & Social Sciences · Entry S0319

Resources

• CSEF project summary ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Behavioural and Social Sciences projects
• CSEF Behavioural and Social Sciences projects
• Every CSEF 2008 project