Reconstructing the detailed geometric structure of a face from a given image
is a key to many computer vision and graphics applications, such as motion
capture and reenactment. The reconstruction task is challenging as human faces
vary extensively when considering expressions, poses, t
本文提供一种基于学习的方法来从单个肖像图像中恢复人头的 3D 几何形状,并且使用参数化 3D 人脸模型来表示头部几何形状以及其他头部区域的深度图,同时使用双目立体匹配方法从具有野外人脸的图像中学习头发和耳朵的几何信息,最后,评估和对比结果表明该方法可以产生高保真的 3D 头部几何形态和头部姿态操作结果。