Model-based Coding of Facial Image Sequences at Varying Illumination Conditions
Peter Eisert, Bernd Girod · 1998
In this paper we describe a model-based algorithm for the estimation of photometric properties in a scene recorded with a video camera. We focus on the coding of head-and-shoulder scenes at very low data-rates of about 1kbit/s. Facial animation parameters [1] specifying facial expressions are estimated from video sequences and transmitted to the decoder. There, the sequence is reconstructed by rendering a 3-D head model that is animated according to the facial parameters. We show in this paper that the quality of the decoded images and the robustness of the motion estimation can be improved by considering photometric e ects. An illumination model based on Lambert re ection of directional colored light is added to the virtual scene and adapted to the current illumination condition. Experimental results show an improvement of about 1.4 dB in PSNR in comparison to simple ambient illumination models. 1