Frequency Perception Network for Camouflaged Object Detection

TL;DR

This work addresses the challenging problem of camouflaged object detection by introducing FPNet, a two-stage framework that jointly leverages RGB and learnable frequency cues. A Frequency Perception Module based on octave convolution decomposes features into high- and low-frequency components to achieve coarse localization, while a Detail-preserving Fine Localization stage with a Correction Fusion Module and cross-layer channel association refines the segmentation using high-resolution cues. The approach yields state-of-the-art or competitive results on COD10K, CAMO, and CHAMELEON, with ablations confirming the contributions of FPM, HRP, and CFM, and frequency analysis highlighting the importance of high-frequency information. The method demonstrates the practical impact of integrating frequency-domain learning into COD and offers a plug-and-play avenue for enhancing camouflage detection in complex scenes.

Abstract

Camouflaged object detection (COD) aims to accurately detect objects hidden in the surrounding environment. However, the existing COD methods mainly locate camouflaged objects in the RGB domain, their performance has not been fully exploited in many challenging scenarios. Considering that the features of the camouflaged object and the background are more discriminative in the frequency domain, we propose a novel learnable and separable frequency perception mechanism driven by the semantic hierarchy in the frequency domain. Our entire network adopts a two-stage model, including a frequency-guided coarse localization stage and a detail-preserving fine localization stage. With the multi-level features extracted by the backbone, we design a flexible frequency perception module based on octave convolution for coarse positioning. Then, we design the correction fusion module to step-by-step integrate the high-level features through the prior-guided correction and cross-layer feature channel association, and finally combine them with the shallow features to achieve the detailed correction of the camouflaged objects. Compared with the currently existing models, our proposed method achieves competitive performance in three popular benchmark datasets both qualitatively and quantitatively.

Figures (9)

• Figure 1: Three challenging camouflaged object detection (COD) scenarios from top to down are with indefinable boundaries, multiple objects, and occluded objects, respectively. The images from left to right are (a) Input image, (b) GT, (c) Ours, (d) SINet-V2 r53, (e) LSR r38.
• Figure 2: The overview of our proposed two-stage network FPNet. The input image is first extracted with multi-level features by a PVT encoder. In the frequency-guided coarse localization stage, we use FPM for frequency-domain feature extraction and generate the coarse COD map $S_1$. Then, in the detail-preserving fine localization stage, the CFM is used to achieve progressively prior-guided correction and fusion across high-level layers. Finally, the first-level high-resolution features are further introduced to refine the boundaries of camouflaged objects and generate the final result $S_{output}$.
• Figure 3: Illustration of frequency-perception module (FPM). Two branches are for high-frequency and low-frequency information learning, respectively.
• Figure 4: The schematic illustration of the correction fusion module (CFM). CFM contains two parts, i.e., prior-guided correction and channel-wise correlation modeling.
• Figure 5: Qualitative results of our proposed FPNet model and some state-of-the-art COD methods. The images from left to right are (a) Input image, (b) GT, (c) Ours, (d) FreNet r39, (e) SINet-V2 r53, (f) PFNet r15, (g) LSR r38, and (h) PraNet r24.