Recursive Head Reconstruction from Multi-View Video Sequences

International audience<p>Face reconstruction from images has been a core topic for the last decades, and is now involved in manyapplications such as identity verification or human–computer interaction. The 3D Morphable Modelintroduced by Blanz and Vetter has been widely used to this end, because its specific 3D modeling offersrobustness to pose variation and adaptability to the specificities of each face.To overcome the limitations of methods using a single image, and since video has become more andmore affordable, we propose a new method which exploits video sequences to consolidate the 3D headshape estimation using successive frames. Based on particle filtering, our algorithm updates the modelestimation at each instant and it is robust to noisy observations. A comparison with the Levenberg–Marquardt global optimization approach on various sets of data shows visual improvements both on poseand shape estimation. Biometric performances confirm this trend with a mean reduction of 10% in termsof False Rejection Rate.