Synthetic-to-Real Camouflaged Object Detection

TL;DR

This paper introduces a novel domain adaptation framework for camouflaged object detection that leverages synthetic data and limited real data to improve detection performance in real-world scenarios.

Contribution

It proposes the CSRDA framework, a student-teacher based method that uses pseudo labeling and recurrent learning to bridge the synthetic and real domains for camouflaged object detection.

Findings

Effective in reducing domain gap between synthetic and real data

Improves detection accuracy with limited real annotations

Demonstrates robustness across various camouflaged object categories

Abstract

Due to the high cost of collection and labeling, there are relatively few datasets for camouflaged object detection (COD). In particular, for certain specialized categories, the available image dataset is insufficiently populated. Synthetic datasets can be utilized to alleviate the problem of limited data to some extent. However, directly training with synthetic datasets compared to real datasets can lead to a degradation in model performance. To tackle this problem, in this work, we investigate a new task, namely Syn-to-Real Camouflaged Object Detection (S2R-COD). In order to improve the model performance in real world scenarios, a set of annotated synthetic camouflaged images and a limited number of unannotated real images must be utilized. We propose the Cycling Syn-to-Real Domain Adaptation Framework (CSRDA), a method based on the student-teacher model. Specially, CSRDA propagates class information from the labeled source domain to the unlabeled target domain through pseudo labeling combined with consistency regularization. Considering that narrowing the intra-domain gap can improve the quality of pseudo labeling, CSRDA utilizes a recurrent learning framework to build an evolving real domain for bridging the source and target domain. Extensive experiments demonstrate the effectiveness of our framework, mitigating the problem of limited data and handcraft annotations in COD. Our code is publicly available at: https://github.com/Muscape/S2R-COD.