Comprehensive Attention Self-Distillation for Weakly-Supervised Object Detection

TL;DR

This paper introduces CASD, a novel self-distillation approach that leverages comprehensive attention across transformations and layers to improve weakly-supervised object detection, achieving state-of-the-art results.

Contribution

The paper proposes a comprehensive attention self-distillation method that enhances feature learning and spatial consistency in WSOD, outperforming existing techniques.

Findings

Achieves new state-of-the-art results on PASCAL VOC and MS-COCO benchmarks.

Effectively balances feature learning among all object instances.

Enforces spatial consistency through self-distillation.

Abstract

Weakly Supervised Object Detection (WSOD) has emerged as an effective tool to train object detectors using only the image-level category labels. However, without object-level labels, WSOD detectors are prone to detect bounding boxes on salient objects, clustered objects and discriminative object parts. Moreover, the image-level category labels do not enforce consistent object detection across different transformations of the same images. To address the above issues, we propose a Comprehensive Attention Self-Distillation (CASD) training approach for WSOD. To balance feature learning among all object instances, CASD computes the comprehensive attention aggregated from multiple transformations and feature layers of the same images. To enforce consistent spatial supervision on objects, CASD conducts self-distillation on the WSOD networks, such that the comprehensive attention is approximated simultaneously by multiple transformations and feature layers of the same images. CASD produces new state-of-the-art WSOD results on standard benchmarks such as PASCAL VOC 2007/2012 and MS-COCO.