Context-aware Attentional Pooling (CAP) for Fine-grained Visual Classification

TL;DR

This paper introduces Context-aware Attentional Pooling (CAP), a novel method that enhances fine-grained visual classification by capturing subtle variations without needing detailed annotations, and demonstrates superior performance across multiple datasets.

Contribution

The paper proposes a new context-aware attentional pooling technique that captures subtle discriminative features and encodes semantic correlations, improving fine-grained classification accuracy.

Findings

Outperforms state-of-the-art methods on six benchmark datasets.

Effectively captures subtle variations without bounding-box annotations.

Compatible with multiple backbone networks.

Abstract

Deep convolutional neural networks (CNNs) have shown a strong ability in mining discriminative object pose and parts information for image recognition. For fine-grained recognition, context-aware rich feature representation of object/scene plays a key role since it exhibits a significant variance in the same subcategory and subtle variance among different subcategories. Finding the subtle variance that fully characterizes the object/scene is not straightforward. To address this, we propose a novel context-aware attentional pooling (CAP) that effectively captures subtle changes via sub-pixel gradients, and learns to attend informative integral regions and their importance in discriminating different subcategories without requiring the bounding-box and/or distinguishable part annotations. We also introduce a novel feature encoding by considering the intrinsic consistency between the informativeness of the integral regions and their spatial structures to capture the semantic correlation among them. Our approach is simple yet extremely effective and can be easily applied on top of a standard classification backbone network. We evaluate our approach using six state-of-the-art (SotA) backbone networks and eight benchmark datasets. Our method significantly outperforms the SotA approaches on six datasets and is very competitive with the remaining two.