Computer vision aided lip movement correction to improve English pronunciation
Abstract
This paper explored the possibility of improving the pronunciation of English as a Foreign Language (EFL) learners by correcting their mouth-lip movement through visual feedback methods. As an EFL learner, I noticed that while learning pronunciation, for some words, the mouth-lip movement (e.g., opening size and duration) will affect the final pronunciation. A prototype system was developed to compare the standard pronunciation mouth shape movement with a user's pronunciation mouth shape movement and give visual feedback to users. Using computer vision technology, the mouth shape movement of standard pronunciation is extracted into moving contours. The user's mouth movement is video recorded. The standard mouth contour movement is overlaid on top of the user's video for comparison. Evaluation found that the computer vision aided lip movement correction method might be effective in improving English pronunciation on a small portion of English words, but could not be generalized to all English words. According to the evaluation data, pronunciation of some words that have obvious mouth shape movement differences might be improved through the proposed computer vision aided lip movement correction method.
Degree
M.S.
Advisors
Chen, Purdue University.
Subject Area
Educational technology|Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.