Reinforced pedestrian attribute recognition with group optimization reward

Abstract

Pedestrian Attribute Recognition (PAR) is a challenging task in intelligent video surveillance. Two key challenges in PAR include complex alignment relations between images and attributes, and imbalanced data distribution. Existing approaches usually formulate PAR as a recognition task. Different from them, this paper addresses it as a decision-making task via a reinforcement learning framework, which is dubbed as Rein-PAR. Specifically, PAR is formulated as a Markov decision process (MDP) to efficiently explore semantic alignments between images and attributes. To alleviate the inter-attribute imbalance problem, we apply an Attribute Grouping Strategy (AGS) by dividing all attributes into subgroups according to their region and category information. Then we employ an agent to recognize each group of attributes, which is trained with Deep Q-learning algorithm. We also propose a Group Optimization Reward (GOR) function to alleviate the intra-attribute imbalance problem. Experimental results on the three benchmark datasets of PETA, RAP and PA100K illustrate the effectiveness and competitiveness of the proposed approach and demonstrate that the application of reinforcement learning to PAR is a valuable research direction. • Formulates Pedestrian Attribute Recognition (PAR) as a Markov decision process. • Applies an Attribute Grouping Strategy to alleviate the attribute imbalance problem. • Develops a novel Group Optimization Reward function. • Extensive experiments demonstrate the effectiveness of the proposed approach.