解决未知类别发现中模型分心问题，提升识别准确率。
A Hidden Stumbling Block in Generalized Category Discovery: Distracted Attention
- 提出注意力聚焦机制，自动剔除无关图像区域的干扰信息。
- 在SimGCD上实现最高15.4%性能提升，计算开销极低。
- 可即插即用，适合希望提升模型专注力的研究者。
广义类别发现（GCD）旨在利用已知类别的标注知识，对包含已知和未知类别的无标签数据进行分类。现有方法虽取得进展，但常忽视一个隐藏障碍：注意力分散。具体而言，模型在处理无标签数据时，不仅关注图像中的关键物体，还会被无关背景区域吸引，导致特征提取不佳。为此，本文提出注意力聚焦（AF）机制，通过剪枝非信息性令牌来增强模型聚焦能力。AF由两个简单有效组件构成：多尺度令牌重要性评估（TIME）与令牌自适应剪枝（TAP），二者级联工作。TIME在多尺度下量化令牌重要性，TAP则利用多尺度得分剪除非信息性令牌。AF为轻量级模块，可无缝集成至现有GCD方法，计算开销极小。将其应用于主流方法SimGCD，性能最高提升15.4%，代码已开源（https://github.com/Afleve/AFGCD）。
原文摘要 · Abstract (English)
Generalized Category Discovery (GCD) aims to classify unlabeled data from both known and unknown categories by leveraging knowledge from labeled known categories. While existing methods have made notable progress, they often overlook a hidden stumbling block in GCD: distracted attention. Specifically, when processing unlabeled data, models tend to focus not only on key objects in the image but also on task-irrelevant background regions, leading to suboptimal feature extraction. To remove this stumbling block, we propose Attention Focusing (AF), an adaptive mechanism designed to sharpen the model's focus by pruning non-informative tokens. AF consists of two simple yet effective components: Token Importance Measurement (TIME) and Token Adaptive Pruning (TAP), working in a cascade. TIME quantifies token importance across multiple scales, while TAP prunes non-informative tokens by utilizing the multi-scale importance scores provided by TIME. AF is a lightweight, plug-and-play module that integrates seamlessly into existing GCD methods with minimal computational overhead. When incorporated into one prominent GCD method, SimGCD, AF achieves up to 15.4% performance improvement over the baseline with minimal computational overhead. The implementation code is provided in https://github.com/Afleve/AFGCD.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。