English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This paper proposes a few-shot method based on Faster R-CNN andrepresentation learning for object detection in aerial images. The twoclassification branches of Faster R-CNN are replaced by prototypical networksfor online adaptation to new classes. These networks produce embeddings vectorsfor each generated box, which are then compared with class prototypes. Thedistance between an embedding and a prototype determines the correspondingclassification score. The resulting networks are trained in an episodic manner.A new detection task is randomly sampled at each epoch, consisting in detectingonly a subset of the classes annotated in the dataset. This training strategyencourages the network to adapt to new classes as it would at test time. Inaddition, several ideas are explored to improve the proposed method such as ahard negative examples mining strategy and self-supervised clustering forbackground objects. The performance of our method is assessed on DOTA, alarge-scale remote sensing images dataset. The experiments conducted provide abroader understanding of the capabilities of representation learning. Ithighlights in particular some intrinsic weaknesses for the few-shot objectdetection task. Finally, some suggestions and perspectives are formulatedaccording to these insights.

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