基于Convnextv2与纹理边缘引导的伪装目标检测

付家瑞,李兆飞,周豪,黄惟

Camouflaged object detection based on Convnextv2 and texture-edge guidance

Jiarui FU,Zhaofei LI,Hao ZHOU,Wei HUANG

表 1 CTEGAFNet与其他11种算法在CAMO和COD10K上的对比结果

Tab.1 Comparison result of CTEGAFNet and other 11 methods in CAMO and COD10K

网络 CAMO-TEST COD10K-TEST Np/106 $ S_{\alpha} $ $ {F_\beta ^{\omega}} $ $ E_{\phi} $ $ \mathrm{MAE} $ $ S_{\alpha} $ $ {F_\beta ^{\omega}} $ $ E_{\phi} $ $ \mathrm{MAE} $ MSCAF 0.873 0.828 0.929 0.046 0.865 0.775 0.927 0.024 28.33 SARNet 0.868 0.828 0.927 0.047 0.864 0.800 0.931 0.024 44.79 FSNet 0.880 0.861 0.933 0.041 0.870 0.810 0.938 0.023 124.53 HitNet 0.844 0.801 0.902 0.057 0.868 0.798 0.932 0.024 24.53 SegMaR 0.815 0.742 0.872 0.071 0.833 0.724 0.895 0.033 68.04 SINet 0.745 0.644 0.829 0.092 0.776 0.631 0.864 0.043 48.95 SINetV2 0.820 0.743 0.882 0.070 0.815 0.680 0.887 0.037 26.98 C2FNet 0.796 0.719 0.864 0.080 0.813 0.686 0.890 0.036 26.30 BGNet 0.812 0.749 0.870 0.073 0.831 0.722 0.901 0.033 74.20 DGNet 0.839 0.769 0.901 0.057 0.822 0.693 0.896 0.033 21.02 ZoomNet 0.820 0.752 0.883 0.066 0.838 0.729 0.893 0.029 32.38 CTEGAFNet 0.893 0.858 0.937 0.037 0.879 0.801 0.933 0.021 92.94