Real-time audiovisual speech capture and motion tracking for speech-driven facial animation
Abstract
Currently, some methods for implementing facial animation systems are based on a direct subphonemic mapping of speech acoustics onto orofacial motion. Although these systems provide all of the necessary components for the detection of facial movements through speech patterns, they are limited by their storage requirements and high-priced equipment. Furthermore, the audiovisual processing can be delayed, taking place after the initial capture and transfer of all required audio and video. The method described in this work focuses on developing a system aimed at achieving realistic facial animations in real-time, incorporating synchronized capture of both audio and facial motions. With data processing taking place at run-time, the system saves only crucial information to file, providing additional storage for larger amounts of data. Results confirm the efficiency of this approach, which provides a fast, accurate and inexpensive tool for speech-driven facial animation.
Description
Due to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to digital@library.tamu.edu, referencing the URI of the item.Includes bibliographical references (leaves 21-22).
Citation
Jablonski, Karl Adam (2003). Real-time audiovisual speech capture and motion tracking for speech-driven facial animation. Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /ETD -TAMU -2003 -Fellows -Thesis -J32.