Recurrent Convolutional Face Alignment

Mainstream direction in face alignment is now dominated by cascaded regression methods. These methods start from an image with an initial shape and build a set of shape increments by computing features with respect to the current shape estimate. These shape increments move the initial shape to the desired location. Despite the advantages of the cascaded methods, they all share two major limitations: (i) shape increments are learned separately from each other in a cascaded manner, (ii) the use of standard generic computer vision features such SIFT, HOG, does not allow these methods to learn problem-specific features. In this work, we propose a novel Recurrent Convolutional Face Alignment method that overcomes these limitations. We frame the standard cascaded alignment problem as a recurrent process and learn all shape increments jointly, by using a recurrent neural network with the gated recurrent unit. Importantly, by combining a convolutional neural network with a recurrent one we alleviate hand-crafted features, widely adopted in the literature and thus allowing the model to learn task-specific features. Moreover, both the convolutional and the recurrent neural networks are learned jointly. Experimental evaluation shows that the proposed method has better performance than the state-of-the-art methods, and further support the importance of learning a single end-to-end model for face alignment.