A Hidden Stumbling Block in Generalized Category Discovery: Distracted Attention
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915400489369600 |
|---|---|
| author | Xu, Qiyu Hu, Zhanxuan Duan, Yu Pei, Ercheng Tai, Yonghang |
| author_facet | Xu, Qiyu Hu, Zhanxuan Duan, Yu Pei, Ercheng Tai, Yonghang |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_14315 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Hidden Stumbling Block in Generalized Category Discovery: Distracted Attention Xu, Qiyu Hu, Zhanxuan Duan, Yu Pei, Ercheng Tai, Yonghang Computer Vision and Pattern Recognition 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. |
| title | A Hidden Stumbling Block in Generalized Category Discovery: Distracted Attention |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2507.14315 |