GridFace: Face Rectification via Learning Local Homography Transformations
This addresses the problem of geometric variations in face recognition for computer vision applications, but it is incremental as it builds on existing homography and learning methods.
The paper tackles facial geometric variations in unconstrained face recognition by proposing GridFace, which rectifies faces using local homography transformations learned end-to-end with a recognition network, resulting in significant performance improvements.
In this paper, we propose a method, called GridFace, to reduce facial geometric variations and improve the recognition performance. Our method rectifies the face by local homography transformations, which are estimated by a face rectification network. To encourage the image generation with canonical views, we apply a regularization based on the natural face distribution. We learn the rectification network and recognition network in an end-to-end manner. Extensive experiments show our method greatly reduces geometric variations, and gains significant improvements in unconstrained face recognition scenarios.