Facial Attributes Classification Using Multi-task Representation Learning

Abstract

This paper presents a new approach for facial attribute classification using a multi-task learning approach. Unlike other approaches that uses hand engineered features, our model learns a shared feature representation that is wellsuited for multiple attribute classification. Learning a joint feature representation enables interaction between different tasks. For learning this shared feature representation we use a Restricted Boltzmann Machine (RBM) based model, enhanced with a factored multi-task component to become Multi-Task Restricted Boltzmann Machine (MT-RBM). Our approach operates directly on faces and facial landmark points to learn a joint feature representation over all the available attributes. We use an iterative learning approach consisting of a bottom-up/top-down pass to learn the shared representation of our multi-task model and at inference we use a bottom-up pass to predict the different tasks. Our approach is not restricted to any type of attributes, however, for this paper we focus only on facial attributes. We evaluate our approach on three publicly available datasets, the Celebrity Faces (CelebA), the Multi-task Facial Landmarks (MTFL), and the ChaLearn challenge dataset. We show superior classification performance improvement over the state-of-the-art.