Spatial-Aware Token for Weakly Supervised Object Localization

TL;DR

This paper introduces a spatial-aware token (SAT) for weakly supervised object localization that improves localization accuracy by focusing on task-specific spatial features and applying novel spatial constraints, achieving state-of-the-art results.

Contribution

The paper proposes a task-specific spatial-aware token and spatial constraints to enhance weakly supervised object localization with transformer models, addressing conflicts between classification and localization.

Findings

Achieves 98.45% GT-known Loc on CUB-200

Achieves 73.13% GT-known Loc on ImageNet

Outperforms state-of-the-art methods, even with limited training data.

Abstract

Weakly supervised object localization (WSOL) is a challenging task aiming to localize objects with only image-level supervision. Recent works apply visual transformer to WSOL and achieve significant success by exploiting the long-range feature dependency in self-attention mechanism. However, existing transformer-based methods synthesize the classification feature maps as the localization map, which leads to optimization conflicts between classification and localization tasks. To address this problem, we propose to learn a task-specific spatial-aware token (SAT) to condition localization in a weakly supervised manner. Specifically, a spatial token is first introduced in the input space to aggregate representations for localization task. Then a spatial aware attention module is constructed, which allows spatial token to generate foreground probabilities of different patches by querying and to extract localization knowledge from the classification task. Besides, for the problem of sparse and unbalanced pixel-level supervision obtained from the image-level label, two spatial constraints, including batch area loss and normalization loss, are designed to compensate and enhance this supervision. Experiments show that the proposed SAT achieves state-of-the-art performance on both CUB-200 and ImageNet, with 98.45% and 73.13% GT-known Loc, respectively. Even under the extreme setting of using only 1 image per class from ImageNet for training, SAT already exceeds the SOTA method by 2.1% GT-known Loc. Code and models are available at https://github.com/wpy1999/SAT.