Facial performance illumination transfer from a single video using interpolation in non-skin region

摘要

This paper proposes a novel video-based method to transfer the illumination from a single reference facial performancenvideo to a target one taken under nearly uniform illumination. We first filter the key frames of the reference and the targetnface videos with an edge-preserving filter. Then, the illumination component of reference key frame is extracted throughndividing the filtered reference key frames by the corresponding filtered target key frames in skin region. The differences innnon-skin region caused by different expressions between the reference and target face may bring about artifacts. Therefore,nwe interpolate the illumination component of the non-skin region by that of the surrounded skin region to ensure the spatialnsmoothness and consistency. After that, the illumination components of key frames are propagated to non-key frames tonensure the temporal consistency between the two adjacent frames. We obtain convincing results by transferring the illuminationneffects of a single reference facial performance video to a target one with the spatial and temporal consistenciesnpreserved. Copyright © 2013 John Wiley & Sons, Ltd.