Facial landmark detection via self-adaption model and multi-task feature learning

Initialization and feature selection are crucial in supervised landmark detection. Mismatching caused by detector error and discrepant initialization is very common in these existing methods. To solve this problem, we proposed a new method, which includes a new initialize model and multitask feature learning, for the robust facial landmark localization. In our new method, firstly, a fast detection will be processed to locate the eyes and mouth, and the initialization model will adapt to the real location according to fast detection. Secondly, a multitask feature learning based Supervised Descent Method (SDM) is adopted on our model to achieve a better performance. For each task, multiple features will be jointly learned and their weights will be calculated during training process. Experiments on three benchmark databases show that our method achieves state-of-art performance.