Cascade Attentive Dropout for Weakly Supervised Object Detection

TL;DR

This paper introduces a cascade attentive dropout method combined with an improved global context module to enhance weakly supervised object detection, effectively addressing part domination and improving detection accuracy on PASCAL VOC 2007.

Contribution

The paper proposes a novel cascade attentive dropout strategy and an enhanced global context module to improve WSOD performance by mitigating part domination issues.

Findings

Achieved 49.8% mAP on PASCAL VOC 2007

Achieved 66.0% CorLoc on PASCAL VOC 2007

Outperformed existing state-of-the-art methods

Abstract

Weakly supervised object detection (WSOD) aims to classify and locate objects with only image-level supervision. Many WSOD approaches adopt multiple instance learning as the initial model, which is prone to converge to the most discriminative object regions while ignoring the whole object, and therefore reduce the model detection performance. In this paper, a novel cascade attentive dropout strategy is proposed to alleviate the part domination problem, together with an improved global context module. We purposely discard attentive elements in both channel and space dimensions, and capture the inter-pixel and inter-channel dependencies to induce the model to better understand the global context. Extensive experiments have been conducted on the challenging PASCAL VOC 2007 benchmarks, which achieve 49.8% mAP and 66.0% CorLoc, outperforming state-of-the-arts.