ATTENTION COUPLENET: FULLY CONVOLUTIONAL ATTENTION COUPLING NETWORK FOR OBJECT DETECTION

Abstract

The field of object detection has made great progress in recent years. Most of these improvements are derived from using a more sophisticated Convolutional Neural Network (CNN). However, in the case of humans, the attention mechanism, global structure information and local details of objects all play an important role for detecting an object. In this paper, we propose a novel fully convolutional network, named as Attention CoupleNet (ACoupleNet), to incorporate the attention-related information and global and local information of objects to improve the detection performance. Specifically, we first design a cascade attention structure to perceive the global scene of the image and generate class-agnostic attention maps. Then the attention maps are encoded into the network to acquire object-aware features. Next, we propose a unique fully convolutional coupling structure to couple global structure and local parts of the object to further formulate a discriminative feature representation. To fully explore the global and local properties, we also design different coupling strategies and normalization ways to make full use of the complementary advantages between the global and local information. Extensive experiments demonstrate the effectiveness of our approach. We achieve state-of-the-art results on all three challenging datasets, i.e. a mAP of 85:7% on VOC07, 84:3% on VOC12, and 35:4% on COCO.

Existing System

RoI-wise subnetwork, RoI pooling layer

Proposed System

A novel cascade attention structure is designed to automatically refine the regions of object and is easy to incorporate with the existing deep networks in a learnable fashion. A fully convolutional coupling structure is proposed to jointly learn the local, global and context information of the object. Different normalization methods and coupling strategies are evaluated to mine the compatibility and complementarity between the global and local information. This information provides the desired output.

Architecture

BLOCK DIAGRAM