The Art of Camouflage: Few-Shot Learning for Animal Detection and Segmentation

TL;DR

This paper introduces a new few-shot learning framework and dataset for detecting and segmenting camouflaged animals, addressing the challenge of limited data and high similarity to backgrounds in natural scenes.

Contribution

The work presents a novel dataset CAMO-FS and a framework FS-CDIS with specialized loss functions for improved camouflaged object detection and segmentation.

Findings

Achieved state-of-the-art results on CAMO-FS dataset.

Demonstrated effectiveness of triplet loss and memory storage in distinguishing camouflaged objects.

Provided a new benchmark for future research in camouflaged object detection.

Abstract

Camouflaged object detection and segmentation is a new and challenging research topic in computer vision. There is a serious issue of lacking data on concealed objects such as camouflaged animals in natural scenes. In this paper, we address the problem of few-shot learning for camouflaged object detection and segmentation. To this end, we first collect a new dataset, CAMO-FS, for the benchmark. As camouflaged instances are challenging to recognize due to their similarity compared to the surroundings, we guide our models to obtain camouflaged features that highly distinguish the instances from the background. In this work, we propose FS-CDIS, a framework to efficiently detect and segment camouflaged instances via two loss functions contributing to the training process. Firstly, the instance triplet loss with the characteristic of differentiating the anchor, which is the mean of all camouflaged foreground points, and the background points are employed to work at the instance level. Secondly, to consolidate the generalization at the class level, we present instance memory storage with the scope of storing camouflaged features of the same category, allowing the model to capture further class-level information during the learning process. The extensive experiments demonstrated that our proposed method achieves state-of-the-art performance on the newly collected dataset. Code is available at https://github.com/danhntd/FS-CDIS.