Distraction-Aware Feature Learning for Human Attribute Recognition via Coarse-to-Fine Attention Mechanism

Abstract

Recently, Human Attribute Recognition (HAR) has become a hot topic due to its scientific challenges and application potentials, where localizing attributes is a crucial stage but not well handled. In this paper, we propose a novel deep learning approach to HAR, namely Distraction-aware HAR (Da-HAR). It enhances deep CNN feature learning by improving attribute localization through a coarse-to-fine attention mechanism. At the coarse step, a self-mask block is built to roughly discriminate and reduce distractions, while at the fine step, a masked attention branch is applied to further eliminate irrelevant regions. Thanks to this mechanism, feature learning is more accurate, especially when heavy occlusions and complex backgrounds exist. Extensive experiments are conducted on the WIDER-Attribute and RAP databases, and state-of-the-art results are achieved, demonstrating the effectiveness of the proposed approach.

8 pages, 5 figures, accepted by AAAI-20 as an oral presentation

References in corpus (6)

• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• A Richly Annotated Dataset for Pedestrian Attribute Recognition
• Localization Guided Learning for Pedestrian Attribute Recognition
• Dynamic Curriculum Learning for Imbalanced Data Classification
• Attribute Aware Pooling for Pedestrian Attribute Recognition
• Weakly-supervised Learning of Mid-level Features for Pedestrian Attribute Recognition and Localization