Contrastive Class-Specific Encoding for Few-Shot Object Detection

Dizhong Lin, Ying Fu, Xin Wang, Shu Hu, Bin Zhu, Qi Song, Xi Wu, Siwei Lyu

Abstract

In this paper, we propose a new few-shot object detection (FSOD) framework that introduces a new contrastive branch to extract the class representation of images, which improves the generalization performance of the detection model for novel classes. Additionally, we investigate the effectiveness of both self-supervised and supervised contrastive losses for class-specific encoding in our framework. Experimental results on the benchmark datasets indicate that our proposed method archives the state-of-the-art performance compared with existing FSOD methods.
BibTeX
@inproceedings{lin2022contrastive,
  author = {Dizhong Lin and Ying Fu and Xin Wang and Shu Hu and Bin Zhu and Qi Song and Xi Wu and Siwei Lyu},
  title = {{Contrastive Class-Specific Encoding for Few-Shot Object Detection}},
  booktitle = {ICME},
  year = {2022},
  doi = {10.1109/icme52920.2022.9859860},
}

Loading related content...

Figure