Unconstrained Face Alignment via Cascaded Compositional Learning

Shizhan Zhu^1,2, Cheng Li², Chen Change Loy^1,3, Xiaoou Tang^1,3

¹Department of Information Engineering, The Chinese University of Hong Kong

²SenseTime Group

³Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences

Abstract

We present a practical approach to address the problem of unconstrained face alignment for a single image. In our unconstrained problem, we need to deal with large shape and appearance variations under extreme head poses and rich shape deformation. To equip cascaded regressors with the capability to handle global shape variation and irregular appearance-shape relation in the unconstrained scenario, we partition the optimisation space into multiple domains of homogeneous descent, and predict a shape as a composition of estimations from multiple domain-specific regressors. With a specially formulated learning objective and a novel tree splitting function, our approach is capable of estimating a robust and meaningful composition. In addition to achieving state-of-the-art accuracy over existing approaches, our framework is also an efficient solution (350 FPS), thanks to the on-the-fly domain exclusion mechanism and the capability of leveraging the fast pixel feature.

Data Release

The revised annotation of full AFLW can be downloaded here. It provides the revised full x-y annotations for ALL landmarks (regardless the visibility), the revised visibility label for ALL landmarks, and the train/test partition benchmarking used in our experiments. Please refer to the 'readme' variable for details.

Code Release

Full codes will be released in July.

Reference

Shizhan Zhu, Cheng Li, Chen Change Loy, Xiaoou Tang. Face alignment by coarse-to-fine shape searching, in Proceedings of Computer Vision and Pattern Recognition (CVPR), 2015. [PDF]