Facial Landmarks Localization Using Cascaded Neural Networks

Shahar Mahpod, Faculty of Engineering, Bar-Ilan University

The accurate localization of facial landmarks is at the core of face analysis tasks, such as face recognition and facial expression analysis, to name a few. In this work we propose a novel localization approach based on a deep learning architecture that utilizes dual cascaded subnetworks with convolutional neural network (CNN) units. The cascaded units of the first subnetwork estimate heatmaps encodings of the landmarks’ locations, while the cascaded units of the second subnetwork receive as input the output of the corresponding heatmap estimation units, and refine them through regression. The proposed scheme is experimentally shown to compare favorably with contemporary state-of-the-art schemes, especially when applied to images depicting challenging localization conditions  .

* PhD research supervised by Prof. Yosi Keller