Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2410.05804 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910899408732160 |
|---|---|
| author | Guo, Mingyi Liu, Yuyang Yan, Zhiyuan Lin, Zongying Peng, Peixi Tian, Yonghong |
| author_facet | Guo, Mingyi Liu, Yuyang Yan, Zhiyuan Lin, Zongying Peng, Peixi Tian, Yonghong |
| contents | Incremental object detection is fundamentally challenged by catastrophic forgetting. A major factor contributing to this issue is background shift, where background categories in sequential tasks may overlap with either previously learned or future unseen classes. To address this, we propose a novel method called Class-Agnostic Shared Attribute Base (CASA) that encourages the model to learn category-agnostic attributes shared across incremental classes. Our approach leverages an LLM to generate candidate textual attributes, selects the most relevant ones based on the current training data, and records their importance in an assignment matrix. For subsequent tasks, the retained attributes are frozen, and new attributes are selected from the remaining candidates, ensuring both knowledge retention and adaptability. Extensive experiments on the COCO dataset demonstrate the state-of-the-art performance of our method. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_05804 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | CASA: Class-Agnostic Shared Attributes in Vision-Language Models for Efficient Incremental Object Detection Guo, Mingyi Liu, Yuyang Yan, Zhiyuan Lin, Zongying Peng, Peixi Tian, Yonghong Computer Vision and Pattern Recognition Incremental object detection is fundamentally challenged by catastrophic forgetting. A major factor contributing to this issue is background shift, where background categories in sequential tasks may overlap with either previously learned or future unseen classes. To address this, we propose a novel method called Class-Agnostic Shared Attribute Base (CASA) that encourages the model to learn category-agnostic attributes shared across incremental classes. Our approach leverages an LLM to generate candidate textual attributes, selects the most relevant ones based on the current training data, and records their importance in an assignment matrix. For subsequent tasks, the retained attributes are frozen, and new attributes are selected from the remaining candidates, ensuring both knowledge retention and adaptability. Extensive experiments on the COCO dataset demonstrate the state-of-the-art performance of our method. |
| title | CASA: Class-Agnostic Shared Attributes in Vision-Language Models for Efficient Incremental Object Detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2410.05804 |