With the rapid development of computer computing power, deep learning plays a more and more important role in the fields of automatic driving, medical research, industrial automation and so on. In order to improve the accuracy of lip-reading recognition, an algorithm based on the model of lip deep learning was proposed in this paper. Binary image of the lip contour motion sequence was projected to the spatio-temporal energy, lip dynamic grayscale was used to reduce noise interference in the recognition process and then lip-reading recognition result was improved by using the excellent characteristics of deep learning ability. The experimental results show that deep learning can obtain the effective characteristics of lip dynamic change from the lip dynamic gray scale and get better recognition results.
CITATION STYLE
Zhu, M. li, Wang, Q. qing, & Luo, J. lin. (2019). Lip-Reading Based on Deep Learning Model. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11345 LNCS, pp. 32–43). Springer Verlag. https://doi.org/10.1007/978-3-662-59351-6_4
Mendeley helps you to discover research relevant for your work.