Deep Reinforced Attention Learning for Quality-Aware Visual Recognition

Abstract

In this paper, we build upon the weakly-supervised generation mechanism of intermediate attention maps in any convolutional neural networks and disclose the effectiveness of attention modules more straightforwardly to fully exploit their potential. Given an existing neural network equipped with arbitrary attention modules, we introduce a meta critic network to evaluate the quality of attention maps in the main network. Due to the discreteness of our designed reward, the proposed learning method is arranged in a reinforcement learning setting, where the attention actors and recurrent critics are alternately optimized to provide instant critique and revision for the temporary attention representation, hence coined as Deep REinforced Attention Learning (DREAL). It could be applied universally to network architectures with different types of attention modules and promotes their expressive ability by maximizing the relative gain of the final recognition performance arising from each individual attention module, as demonstrated by extensive experiments on both category and instance recognition benchmarks.

A related work (DIANet in AAAI-20) should be discussed and cited

References in corpus (9)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Neural Architecture Search with Reinforcement Learning
• On the Convergence of Adam and Beyond
• AMC: AutoML for Model Compression and Acceleration on Mobile Devices
• Recurrent Models of Visual Attention
• Gather-Excite: Exploiting Feature Context in Convolutional Neural Networks
• Designing Neural Network Architectures using Reinforcement Learning
• Continuous Deep Q-Learning with Model-based Acceleration
• N2N Learning: Network to Network Compression via Policy Gradient Reinforcement Learning