A style transfer-based augmentation approach for detecting military camouflaged object
DOI:
https://doi.org/10.54939/1859-1043.j.mst.CSCE8.2024.44-54Keywords:
Style transfer; Military camouflaged object detection; Data augmentation.Abstract
Detecting camouflaged objects in military environments is particularly challenging due to the deliberate blending of targets with their surroundings. Despite advancements in deep learning, the limited availability of training data remains a significant obstacle. To address this issue, this paper proposes a novel data augmentation approach that combines multiple style transfer models to transform the style of training images into diverse reference styles. This enriches the training data by simulating various environmental textures and patterns. Structural Similarity (SSIM) is used as an evaluation metric to select style-transferred images that preserve the best structural similarity, enabling detection models to more effectively learn the distinguishing features of camouflaged objects. Extensive experiments with different style transfer methods and SSIM thresholds show that our augmentation approach significantly enhances the accuracy of state-of-the-art detection algorithms. This approach has the potential to improve object detection in military operations, increasing the reliability and precision of automated surveillance systems.
References
[1]. An, Jie, Siyu Huang, Yibing Song, Dejing Dou, Wei Liu and Jiebo Luo. “ArtFlow: Unbiased Image Style Transfer via Reversible Neural Flows.” 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR): 862-871, (2021). DOI: https://doi.org/10.1109/CVPR46437.2021.00092
[2]. Du, Siyu, Cuili Yao, Yuqiu Kong and Yicheng Yang. “BANet: Camouflaged Object De-tection Based on Boundary Guidance and Multiple Attention Mechanisms.” 2023 9th Annual International Conference on Network and Information Systems for Computers (ICNISC): 464-469, (2023). DOI: https://doi.org/10.1109/ICNISC60562.2023.00027
[3]. Eker, Thijs A., Friso G. Heslinga, Luca Ballan, Richard J. M. den Hollander and Klamer Schutte. “The effect of simulation variety on a deep learning-based military vehicle detector.” Security + Defence, (2023). DOI: https://doi.org/10.1117/12.2679808
[4]. Fan, Weichen, Jinghuan Chen and Ziwei Liu. “Hierarchy Flow For High-Fidelity Image-to-Image Translation.” ArXiv abs/2308.06909, (2023).
[5]. Guo, Chen, Fabian Benitez-Quiroz, Qianli Feng and Aleix M. Martínez. “Structural Similarity: When to Use Deep Generative Models on Imbalanced Image Dataset Aug-mentation”. (2023).
[6]. Huang, Bin, Ziyue Xu, Shing-Chow Chan, Zhong Liu, Huiying Wen, Chao Hou, Qicai Huang, Meiqin Jiang, Changfeng Dong, Jie Zeng, Ruhai Zou, Bin Huang, Xin Chen and Shuo Li. “A Style Transfer-Based Augmentation Framework for Improving Seg-mentation and Classification Performance Across Different Sources in Ultrasound Im-ages.” International Conference on Medical Image Computing and Computer-Assisted Intervention, (2023). DOI: https://doi.org/10.1007/978-3-031-43987-2_5
[7]. Huang, Xun and Serge J. Belongie. “Arbitrary Style Transfer in Real-Time with Adaptive Instance Normalization.” 2017 IEEE International Conference on Computer Vision (ICCV): 1510-1519, (2017). DOI: https://doi.org/10.1109/ICCV.2017.167
[8]. Hwang, Kyo-Seong and Jungmok Ma. “Military camouflaged object detection with deep learning using dataset development and combination.” The Journal of Defense Modeling and Simulation: Applications, Methodology, Technology (2024). DOI: https://doi.org/10.1177/15485129241233299
[9]. Kanchanadevi, st K. and G. K. Sandhia. “A Neural Style Transfer Network for Image Augmentation of Corn Leaves using Convoluvational Neural Netwok.” 2023 Interna-tional Conference on Integrated Intelligence and Communication Systems (ICIICS): 1-5, (2023). DOI: https://doi.org/10.1109/ICIICS59993.2023.10421717
[10]. Lee, Hyeongkeun, Kyungmin Lee, Hunmin Yang and Se-Yoon Oh. “Applying FastPhotoStyle to Synthetic Data for Military Vehicle Detection.” 2020 20th Interna-tional Conference on Control, Automation and Systems (ICCAS): 137-140, (2020). DOI: https://doi.org/10.23919/ICCAS50221.2020.9268331
[11]. Li, Yijun, Chen Fang, Jimei Yang, Zhaowen Wang, Xin Lu and Ming-Hsuan Yang. “Universal Style Transfer via Feature Transforms.” Neural Information Processing Systems (2017).
[12]. Liu, Mao Zhen and Xiaoguang Di. “Extraordinary MHNet: Military high-level camouflage object detection network and dataset.” Neurocomputing 549: 126466, (2023). DOI: https://doi.org/10.1016/j.neucom.2023.126466
[13]. Maruyama, Sho. “Properties of the SSIM metric in medical image assessment: corre-spondence between measurements and the spatial frequency spectrum.” Physical and Engineering Sciences in Medicine 46: 1131-1141 (2023). DOI: https://doi.org/10.1007/s13246-023-01280-1
[14]. Peng, Yiguo, Jianzhong Wang, Zibo Yu, Yu You and Yong Sun. “Camouflage soldier object detection network based on the attention mechanism and pyramidal feature shrinking.” Multimedia Tools and Applications, (2024). DOI: https://doi.org/10.1007/s11042-024-18618-w
[15]. Rao, Shuzhen, Jun Huang and Zengming Tang. “Exploiting Style Transfer-based Task Augmentation for Cross-Domain Few-Shot Learning.” ArXiv abs/2301.07927 (2023). DOI: https://doi.org/10.2139/ssrn.4424628
[16]. Raveendran, Ritin, Aviral Kumar Singh and M RajeshKumar. “Complex Wavelet SSIM based Image Data Augmentation.” ArXiv abs/2007.05853 (2020).
[17]. Wang, Zhou, Alan Conrad Bovik, Hamid R. Sheikh and Eero P. Simoncelli. “Image quality assessment: from error visibility to structural similarity.” IEEE Transactions on Image Processing 13: 600-612, (2004). DOI: https://doi.org/10.1109/TIP.2003.819861
[18]. Winderlich, Kristina, Clementine Dalelane and Andreas Walter. “Classification of syn-optic circulation patterns with a two-stage clustering algorithm using the structural similarity index metric (SSIM).” (2022). DOI: https://doi.org/10.5194/esd-2022-29
[19]. Yan, Zhancai and Liu Yaqiu. “Image confusion algorithm using adaptive threshold SSIM based on Contourlet transformation.” Australian Journal of Mechanical Engi-neering 16: 23 – 29, (2018). DOI: https://doi.org/10.1080/1448837X.2018.1545468
[20]. Zheng, Xu, Tejo Chalasani, Koustav Ghosal, Sebastian Lutz and Aljosa Smolic. “STaDA: Style Transfer as Data Augmentation.” VISIGRAPP (2019). DOI: https://doi.org/10.5220/0007353400002108
[21]. Zhou, Yubeibei, Jiulu Gong, Weijian Lu, Naiwei Gu, Kuiqi Chong and Zepeng Wang. “An Object Detection Algorithm for Military Vehicles Based on Image Style Transfer and Domain Adversarial Learning.” 2022 IEEE International Conference on Unmanned Systems (ICUS): 1059-1064, (2022). DOI: https://doi.org/10.1109/ICUS55513.2022.9987190
[22]. Zhu, Jun-Yan, Taesung Park, Phillip Isola and Alexei A. Efros. “Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks.” 2017 IEEE International Conference on Computer Vision (ICCV): 2242-2251, (2017). DOI: https://doi.org/10.1109/ICCV.2017.244
[23]. Yolov8 is available at https://github.com/ultralytics/ultralytics.
