-
摘要:
伪装目标检测是一项新兴的视觉检测任务，旨在识别出完美隐藏在周围环境中的伪装目标，在多个领域中具有广泛应用。针对当前伪装目标检测算法无法准确、完整地识别目标结构和边界的问题，基于人类在观察伪装图像时的视觉感知过程，设计了一种生物启发式框架，并命名为定位和细化网络（PRNet）。利用Res2Net提取图像的原始特征，从多层级信息中挖掘目标的边缘线索；特别设计特征增强模块，在丰富全局上下文信息的同时能够扩大感受野；定位模块利用双注意力机制从通道和空间2个维度来定位目标的大致位置；细化模块同时关注前景和背景中的目标线索，利用多类型信息进一步细化目标的结构和边缘。在3个广泛使用的伪装目标检测基准数据集上的大量实验结果表明，所提网络的整体性能明显优于14种比较算法，在多种复杂场景中表现优异。
Abstract:Camouflaged object identification is a new visual detection job that has many applications in several fields. Its goal is to identify camouflaged targets that are completely disguised in their environment. To address the problem of current camouflaged object detection algorithms failing to accurately and completely identify the object's structure and boundaries, this paper designs a bio-heuristic framework based on the human visual perception process when observing camouflaged images, and names its Positioning and Refinement Network (PRNet). Res2Net is used to extract the original features of the image and mine the edge cues of the target from multi-level information. A feature enhancement module is specially designed to expand the perceptual field while enriching the global contextual information. Then, the positioning module utilizes the dual-attention mechanism to locate the approximate position of the target from both channel and spatial dimensions. Lastly, the refinement module leverages multi-type information to further improve the target structure and edges by concentrating on target cues in both the foreground and background. Extensive experimental results on three widely used benchmark datasets for camouflage target detection demonstrate that the overall performance of the proposed network significantly outperforms 14 comparative algorithms and performs well in a variety of complex scenarios.
-
图 6 不同模型的PR曲线和F-measure曲线
Figure 6. PR curves and F-measure curves for different models
表 1 消融实验定量结果（“↑”/“↓”分别表示值越大/小越好）
Table 1. Quantitative results of ablation experiments（“↑”/“↓” indicates that larger/smaller is better.）
模型 CAMO COD10K NC4K S↑ E↑ F↑ M↓ S↑ E↑ F↑ M↓ S↑ E↑ F↑ M↓ Baseline 0.533 0.523 0.304 0.197 0.584 0.583 0.307 0.105 0.585 0.572 0.367 0.153 Baseline+FEM 0.758 0.81 0.704 0.087 0.781 0.849 0.690 0.041 0.812 0.868 0.767 0.058 Baseline+PM 0.639 0.664 0.507 0.145 0.666 0.709 0.481 0.078 0.69 0.719 0.575 0.109 Baseline+RM 0.778 0.834 0.741 0.085 0.802 0.871 0.709 0.037 0.830 0.885 0.794 0.052 PRNet 0.837 0.893 0.799 0.051 0.834 0.904 0.756 0.032 0.875 0.897 0.820 0.041 
下载: 导出CSV
表 2 不同模型的定量检测结果
Table 2. Quantitative detection results of different models
方法 CAMO COD10K NC4K S↑ E↑ F↑ M↓ S↑ E↑ F↑ M↓ S↑ E↑ F↑ M↓ SINetV2[13] 0.815 0.870 0.783 0.074 0.813 0.886 0.713 0.037 0.845 0.901 0.802 0.048 JCSOD[32] 0.767 0.810 0.729 0.086 0.800 0.872 0.718 0.036 0.835 0.886 0.805 0.049 Rank-Net[27] 0.785 0.842 0.736 0.081 0.786 0.863 0.672 0.043 0.823 0.883 0.771 0.054 MGL[33] 0.570 0.499 0.302 0.182 0.635 0.584 0.341 0.111 0.662 0.596 0.428 0.136 PFNet[12] 0.776 0.832 0.738 0.087 0.799 0.875 0.700 0.039 0.828 0.886 0.786 0.053 OCENet[34] 0.775 0.818 0.709 0.092 0.789 0.854 0.680 0.044 0.822 0.871 0.766 0.057 ERRNet[22] 0.767 0.801 0.671 0.104 0.744 0.801 0.578 0.063 0.792 0.833 0.692 0.077 BgNet[23] 0.652 0.684 0.534 0.137 0.657 0.718 0.475 0.082 0.699 0.749 0.594 0.101 ZoomNet[14] 0.797 0.842 0.765 0.076 0.819 0.864 0.742 0.032 0.838 0.878 0.803 0.047 BSANet[20] 0.800 0.852 0.770 0.076 0.812 0.876 0.733 0.035 0.836 0.887 0.803 0.050 C2FNet[35] 0.768 0.823 0.726 0.089 0.808 0.883 0.724 0.036 0.834 0.890 0.798 0.049 FAPNet[16] 0.798 0.849 0.758 0.082 0.820 0.885 0.727 0.036 0.849 0.898 0.805 0.048 MFFN[36] 0.790 0.837 0.752 0.081 0.826 0.872 0.753 0.037 0.837 0.877 0.800 0.052 ASBI[37] 0.821 0.874 0.782 0.073 0.829 0.896 0.742 0.033 0.857 0.907 0.818 0.044 本文 0.837 0.893 0.799 0.051 0.834 0.904 0.756 0.032 0.875 0.897 0.820 0.041
下载: 导出CSV
-
[1] CHUDZIK P, MITCHELL A, ALKASEEM M, et al. Mobile real-time grasshopper detection and data aggregation framework[J]. Scientific Reports, 2020, 10(1): 1150. doi: 10.1038/s41598-020-57674-8 [2] CHEN S M, XIONG J T, JIAO J M, et al. Citrus fruits maturity detection in natural environments based on convolutional neural networks and visual saliency map[J]. Precision Agriculture, 2022, 23(5): 1515-1531. doi: 10.1007/s11119-022-09895-2 [3] LIN Z S, YE H X, ZHAN B, et al. An efficient network for surface defect detection[J]. Applied Sciences, 2020, 10(17): 6085. doi: 10.3390/app10176085 [4] JI G P, CHOU Y C, FAN D P, et al. Progressively normalized self-attention network for video polyp segmentation[C]//Proceedings of the Medical Image Computing and Computer Assisted Intervention – MICCAI 2021. Cham: Springer, 2021: 142-152. [5] WU Y H, GAO S H, MEI J, et al. JCS: an explainable COVID-19 diagnosis system by joint classification and segmentation[J]. IEEE Transactions on Image Processing, 2021, 30: 3113-3126. doi: 10.1109/TIP.2021.3058783 [6] YANG Z Y, DAI Q H, ZHANG J S. Visual perception driven collage synthesis[J]. Computational Visual Media, 2022, 8(1): 79-91. doi: 10.1007/s41095-021-0226-8 [7] HORVÁTH G, PERESZLÉNYI Á, ÅKESSON S, et al. Striped bodypainting protects against horseflies[J]. Royal Society Open Science, 2019, 6(1): 181325. doi: 10.1098/rsos.181325 [8] MERILAITA S, SCOTT-SAMUEL N E, CUTHILL I C. How camouflage works[J]. Philosophical Transactions of the Royal Society B: Biological Sciences, 2017, 372(1724): 20160341. doi: 10.1098/rstb.2016.0341 [9] SENGOTTUVELAN P, WAHI A, SHANMUGAM A. Performance of decamouflaging through exploratory image analysis[C]// 2008 First International Conference on Emerging Trends in Engineering and Technology. Piscataway: IEEE Press, 2008: 6-10. [10] YIN J Q, HAN Y B, HOU W D, et al. Detection of the mobile object with camouflage color under dynamic background based on optical flow[J]. Procedia Engineering, 2011, 15: 2201-2205. doi: 10.1016/j.proeng.2011.08.412 [11] FAN D P, JI G P, SUN G L, et al. Camouflaged object detection[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2020: 2774-2784. [12] MEI H Y, JI G P, WEI Z Q, et al. Camouflaged object segmentation with distraction mining[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2021: 8768-8777. [13] FAN D P, JI G P, CHENG M M, et al. Concealed object detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(10): 6024-6042. doi: 10.1109/TPAMI.2021.3085766 [14] PANG Y W, ZHAO X Q, XIANG T Z, et al. Zoom in and out: a mixed-scale triplet network for camouflaged object detection[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2022: 2150-2160. [15] REN J J, HU X W, ZHU L, et al. Deep texture-aware features for camouflaged object detection[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33(3): 1157-1167. doi: 10.1109/TCSVT.2021.3126591 [16] ZHOU T, ZHOU Y, GONG C, et al. Feature aggregation and propagation network for camouflaged object detection[J]. IEEE Transactions on Image Processing, 2022, 31: 7036-7047. doi: 10.1109/TIP.2022.3217695 [17] GAO S H, CHENG M M, ZHAO K, et al. Res2Net: a new multi-scale backbone architecture[J]. IEEE Trans Pattern Anal Mach Intell, 2021, 43(2): 652-662. doi: 10.1109/TPAMI.2019.2938758 [18] SUN Y J, WANG S, CHEN C, et al. Boundary-guided camouflaged object detection[EB/OL]. (2022-07-02)[2023-07-25]. http://doi.org/10.48550/arXiv.2207.00794. [19] WANG K, BI H B, ZHANG Y, et al. D2C-Net: a dual-branch, dual-guidance and cross-refine network for camouflaged object detection[J]. IEEE Transactions on Industrial Electronics, 2022, 69(5): 5364-5374. doi: 10.1109/TIE.2021.3078379 [20] ZHU H W, LI P, XIE H R, et al. I can find you! boundary-guided separated attention network for camouflaged object detection[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2022, 36(3): 3608-3616. doi: 10.1609/aaai.v36i3.20273 [21] QIN X B, ZHANG Z C, HUANG C Y, et al. BASNet: boundary-aware salient object detection[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2019: 7471-7481. [22] JI G P, ZHU L, ZHUGE M C, et al. Fast camouflaged object detection via edge-based reversible re-calibration network[J]. Pattern Recognition, 2022, 123: 108414. doi: 10.1016/j.patcog.2021.108414 [23] CHEN T Y, XIAO J, HU X G, et al. Boundary-guided network for camouflaged object detection[J]. Knowledge-Based Systems, 2022, 248: 108901. doi: 10.1016/j.knosys.2022.108901 [24] WEI J, WANG S H, HUANG Q M. F³Net: fusion, feedback and focus for salient object detection[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34(7): 12321-12328. doi: 10.1609/aaai.v34i07.6916 [25] XIE E Z, WANG W J, WANG W H, et al. Segmenting transparent objects in the wild[C]//Proceedings of the Computer Vision – ECCV 2020. Berlin: Springer, 2020: 696-711. [26] LE T N, NGUYEN T V, NIE Z L, et al. Anabranch network for camouflaged object segmentation[J]. Computer Vision and Image Understanding, 2019, 184: 45-56. doi: 10.1016/j.cviu.2019.04.006 [27] LV Y Q, ZHANG J, DAI Y C, et al. Simultaneously localize, segment and rank the camouflaged objects[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2021: 11586-11596. [28] ACHANTA R, HEMAMI S, ESTRADA F, et al. Frequency-tuned salient region detection[C]//Proceedings of the 2009 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2009: 1597-1604. [29] CHENG M M, FAN D P. Structure-measure: a new way to evaluate foreground maps[J]. International Journal of Computer Vision, 2021, 129(9): 2622-2638. doi: 10.1007/s11263-021-01490-8 [30] FAN D P, JI G P, QIN X B, et al. Cognitive vision inspired object segmentation metric and loss function[J]. Scientia Sinica Informationis, 2021, 51(9): 1475. doi: 10.1360/SSI-2020-0370 [31] PERAZZI F, KRÄHENBÜHL P, PRITCH Y, et al. Saliency filters: Contrast based filtering for salient region detection[C]//Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2012: 733-740. [32] LI A X, ZHANG J, LV Y Q, et al. Uncertainty-aware joint salient object and camouflaged object detection[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2021: 10066-10076. [33] ZHAI Q, LI X, YANG F, et al. Mutual graph learning for camouflaged object detection[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2021: 12992-13002. [34] LIU J W, ZHANG J, BARNES N. Confidence-aware learning for camouflaged object detection[EB/OL]. (2021-06-22)[2023-07-13]. http://doi.org/10.48550/arXiv.2106.11641. [35] CHEN G, LIU S J, SUN Y J, et al. Camouflaged object detection via context-aware cross-level fusion[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2022, 32(10): 6981-6993. doi: 10.1109/TCSVT.2022.3178173 [36] ZHENG D H, ZHENG X C, YANG L T, et al. MFFN: multi-view feature fusion network for camouflaged object detection[EB/OL]. (2021-06-22)[2023-07-15]. http://doi.org/10.48550/arXiv.2210.06361. [37] ZHANG Q, SUN X X, CHEN Y R, et al. Attention-induced semantic and boundary interaction network for camouflaged object detection[J]. Computer Vision and Image Understanding, 2023, 233(8): 103719. doi: 10.1016/j.cviu.2023.103719 -
下载:
点击查看大图
计量
- 文章访问数: 273
- HTML全文浏览量: 142
- PDF下载量: 19
- 被引次数: 0
