Experience feedback using Representation Learning for Few-Shot Object Detection on Aerial Images

TL;DR

This paper introduces a few-shot object detection approach for aerial images using Faster R-CNN with prototypical networks, enabling online adaptation to new classes through representation learning and episodic training.

Contribution

It presents a novel few-shot detection method combining Faster R-CNN and prototypical networks, with training strategies tailored for aerial image datasets.

Findings

Effective adaptation to new classes demonstrated on DOTA dataset

Identified intrinsic weaknesses in few-shot object detection tasks

Proposed strategies improve detection performance in limited data scenarios

Abstract

This paper proposes a few-shot method based on Faster R-CNN and representation learning for object detection in aerial images. The two classification branches of Faster R-CNN are replaced by prototypical networks for online adaptation to new classes. These networks produce embeddings vectors for each generated box, which are then compared with class prototypes. The distance between an embedding and a prototype determines the corresponding classification score. The resulting networks are trained in an episodic manner. A new detection task is randomly sampled at each epoch, consisting in detecting only a subset of the classes annotated in the dataset. This training strategy encourages the network to adapt to new classes as it would at test time. In addition, several ideas are explored to improve the proposed method such as a hard negative examples mining strategy and self-supervised clustering for background objects. The performance of our method is assessed on DOTA, a large-scale remote sensing images dataset. The experiments conducted provide a broader understanding of the capabilities of representation learning. It highlights in particular some intrinsic weaknesses for the few-shot object detection task. Finally, some suggestions and perspectives are formulated according to these insights.