摘要: An algorithm for speaker's lip segmentation and features extraction is presented. A color video sequence of face acquired, under natural lighting conditions without any particular make-up. First, a logarithmic transform performed from the RGB to HI (hue, intensity) space. Second, statistical approach using Markov random field modeling determines red hue prevailing region motion in spatiotemporal neighborhood. Third, final label used extract ROI (region interest) geometrical features.
sci-hub.se 参考文章(5)
F. Luthon, M. Lievin, Lip motion automatic detection scandinavian conference on image analysis. pp. 253- 260 ,(1997)
Stuart Geman, Donald Geman, Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. PAMI-6, pp. 721- 741 ,(1984) , 10.1109/TPAMI.1984.4767596
Michel Jourlin, Jean-Charles Pinoli, Image dynamic range enhancement and stabilization in the context of the logarithmic image processing model Signal Processing. ,vol. 41, pp. 225- 237 ,(1995) , 10.1016/0165-1684(94)00102-6
Speechreading by Humans and Machines Springer Berlin Heidelberg. ,(1996) , 10.1007/978-3-662-13015-5
C. Benoit, Tayeb Mohamadi, T. Lallouache, C. Abry, A set of French visemes for visual speech synthesis Les Cahiers de l'ICP. Rapport de recherche. pp. 113- 129 ,(1994)