Photorealistic adaptation and interpolation of facial expressions using HMMS and AAMS for audio-visual speech synthesis | IEEE Conference Publication | IEEE Xplore