Learning and Transferring Multi-task Deep Representation for Face Alignment.
Zhanpeng Zhang, Ping Luo, Chen Change Loy, Xiaoou Tang · 2014
Abstract. Facial landmark detection of face alignment has long been impeded by the problems of occlusion and pose variation. Instead of treating the detection task as a single and independent problem, we in-vestigate the possibility of improving detection robustness through multi-task learning. Specifically, we wish to optimize facial landmark detection together with heterogeneous but subtly correlated tasks, e.g. head pose estimation and facial attribute inference. This is non-trivial since dif-ferent tasks have different learning difficulties and convergence rates. To address this problem, we formulate a novel tasks-constrained deep model, with task-wise early stopping to facilitate learning convergence. Extensive evaluations show that the proposed task-constrained learning (i) outperforms existing methods, especially in dealing with faces with severe occlusion and pose variation, and (ii) reduces model complexity drastically compared to the state-of-the-art method based on cascaded deep model [29]. In this technical report, we extend the method presented in our ECCV 2014 [39] paper to handle more landmark points (68 points instead of 5 major facial points) without either redesigning the deep model or in-volving significant increase in run time cost. This is made possible by transferring the learned 5-point model to the desired facial landmark configuration, through model fine-tuning with dense landmark annota-tions. Our new model achieves the state-of-the-art result on the 300-W benchmark dataset (mean error of 9.15 % on the challenging IBUG sub-set). 1