DEEP MULTI-TASK LEARNING FOR FACE AND HUMAN ANALYSIS

Abstract

In this thesis, we use multi-task learning methods to solve face and human analysis tasks. We design multi-task learning models to learn multiple face and human analysis tasks. We demonstrate that deep multi-task learning can be used to perform the face attribute classification task and up to 40 face attributes can be classified simultaneously with one model. We also demonstrate that two challenging pixel-level classification tasks, i.e. human parsing and human instance segmentation, can be addressed within one model to achieve fine-grained human analysis in images. Built on the commonly used deep multi-task learning architecture, we further model the interactions and relations among tasks in multi-task learning. For task interaction modelling, we propose an integrated face analytics network to explicitly enable the interactions of multiple tasks. For task relation modelling, we propose a task relation network to leverage the similarities between tasks in multi-task learning.