Features Split and Aggregation Network for Camouflaged Object Detection
Higher standards have been proposed for detection systems since camouflaged objects are not distinct enough, making it possible to ignore the difference between their background and foreground. In this paper, we present a new framework for Camouflaged Object Detection (COD) named FSANet, which consi...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2024-01-01
|
Series: | Journal of Imaging |
Subjects: | |
Online Access: | https://www.mdpi.com/2313-433X/10/1/24 |
_version_ | 1797343333121523712 |
---|---|
author | Zejin Zhang Tao Wang Jian Wang Yao Sun |
author_facet | Zejin Zhang Tao Wang Jian Wang Yao Sun |
author_sort | Zejin Zhang |
collection | DOAJ |
description | Higher standards have been proposed for detection systems since camouflaged objects are not distinct enough, making it possible to ignore the difference between their background and foreground. In this paper, we present a new framework for Camouflaged Object Detection (COD) named FSANet, which consists mainly of three operations: spatial detail mining (SDM), cross-scale feature combination (CFC), and hierarchical feature aggregation decoder (HFAD). The framework simulates the three-stage detection process of the human visual mechanism when observing a camouflaged scene. Specifically, we have extracted five feature layers using the backbone and divided them into two parts with the second layer as the boundary. The SDM module simulates the human cursory inspection of the camouflaged objects to gather spatial details (such as edge, texture, etc.) and fuses the features to create a cursory impression. The CFC module is used to observe high-level features from various viewing angles and extracts the same features by thoroughly filtering features of various levels. We also design side-join multiplication in the CFC module to avoid detail distortion and use feature element-wise multiplication to filter out noise. Finally, we construct an HFAD module to deeply mine effective features from these two stages, direct the fusion of low-level features using high-level semantic knowledge, and improve the camouflage map using hierarchical cascade technology. Compared to the nineteen deep-learning-based methods in terms of seven widely used metrics, our proposed framework has clear advantages on four public COD datasets, demonstrating the effectiveness and superiority of our model. |
first_indexed | 2024-03-08T10:46:06Z |
format | Article |
id | doaj.art-4522f2c2f0ee435fb10c3755bd504000 |
institution | Directory Open Access Journal |
issn | 2313-433X |
language | English |
last_indexed | 2024-03-08T10:46:06Z |
publishDate | 2024-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Journal of Imaging |
spelling | doaj.art-4522f2c2f0ee435fb10c3755bd5040002024-01-26T17:11:29ZengMDPI AGJournal of Imaging2313-433X2024-01-011012410.3390/jimaging10010024Features Split and Aggregation Network for Camouflaged Object DetectionZejin Zhang0Tao Wang1Jian Wang2Yao Sun3HDU-ITMO Joint Institute, Hangzhou Dianzi University, Hangzhou 310018, ChinaHDU-ITMO Joint Institute, Hangzhou Dianzi University, Hangzhou 310018, ChinaHDU-ITMO Joint Institute, Hangzhou Dianzi University, Hangzhou 310018, ChinaHDU-ITMO Joint Institute, Hangzhou Dianzi University, Hangzhou 310018, ChinaHigher standards have been proposed for detection systems since camouflaged objects are not distinct enough, making it possible to ignore the difference between their background and foreground. In this paper, we present a new framework for Camouflaged Object Detection (COD) named FSANet, which consists mainly of three operations: spatial detail mining (SDM), cross-scale feature combination (CFC), and hierarchical feature aggregation decoder (HFAD). The framework simulates the three-stage detection process of the human visual mechanism when observing a camouflaged scene. Specifically, we have extracted five feature layers using the backbone and divided them into two parts with the second layer as the boundary. The SDM module simulates the human cursory inspection of the camouflaged objects to gather spatial details (such as edge, texture, etc.) and fuses the features to create a cursory impression. The CFC module is used to observe high-level features from various viewing angles and extracts the same features by thoroughly filtering features of various levels. We also design side-join multiplication in the CFC module to avoid detail distortion and use feature element-wise multiplication to filter out noise. Finally, we construct an HFAD module to deeply mine effective features from these two stages, direct the fusion of low-level features using high-level semantic knowledge, and improve the camouflage map using hierarchical cascade technology. Compared to the nineteen deep-learning-based methods in terms of seven widely used metrics, our proposed framework has clear advantages on four public COD datasets, demonstrating the effectiveness and superiority of our model.https://www.mdpi.com/2313-433X/10/1/24bio-inspired networkcontext-aware featuresmulti-scale featurescamouflaged object detection |
spellingShingle | Zejin Zhang Tao Wang Jian Wang Yao Sun Features Split and Aggregation Network for Camouflaged Object Detection Journal of Imaging bio-inspired network context-aware features multi-scale features camouflaged object detection |
title | Features Split and Aggregation Network for Camouflaged Object Detection |
title_full | Features Split and Aggregation Network for Camouflaged Object Detection |
title_fullStr | Features Split and Aggregation Network for Camouflaged Object Detection |
title_full_unstemmed | Features Split and Aggregation Network for Camouflaged Object Detection |
title_short | Features Split and Aggregation Network for Camouflaged Object Detection |
title_sort | features split and aggregation network for camouflaged object detection |
topic | bio-inspired network context-aware features multi-scale features camouflaged object detection |
url | https://www.mdpi.com/2313-433X/10/1/24 |
work_keys_str_mv | AT zejinzhang featuressplitandaggregationnetworkforcamouflagedobjectdetection AT taowang featuressplitandaggregationnetworkforcamouflagedobjectdetection AT jianwang featuressplitandaggregationnetworkforcamouflagedobjectdetection AT yaosun featuressplitandaggregationnetworkforcamouflagedobjectdetection |