YoloV8改进策略：卷积篇|Kan行天下之ReluKan

AI智韵

于 2024-07-20 09:24:26 发布

阅读量186

点赞数

文章标签： YOLO

不能复制，转发

本文链接：https://blog.csdn.net/m0_47867638/article/details/140565755

版权

YOLOv8l summary: 598 layers, 51190656 parameters, 0 gradients, 158.9 GFLOPs
                 Class     Images  Instances      Box(P          R      mAP50  mAP50-95): 100%|██████████| 15/15 [00:01<00:00,  9.41it/s]
                   all        230       1412      0.964      0.975      0.991      0.752
                   c17         40        131       0.98      0.992      0.995      0.826
                    c5         19         68      0.985      0.983      0.994      0.841
            helicopter         13         43      0.949          1      0.984      0.634
                  c130         20         85      0.998      0.988      0.995       0.67
                   f16         11         57      0.982      0.948      0.988      0.681
                    b2          2          2      0.871          1      0.995      0.676
                 other         13         86      0.959      0.953      0.978      0.573
                   b52         21         70      0.985      0.969      0.984      0.841
                  kc10         12         62          1      0.976      0.989      0.859
               command         12         40      0.994          1      0.995      0.824
                   f15         21        123          1      0.962      0.995      0.686
                 kc135         24         91      0.984      0.989       0.99      0.713
                   a10          4         27      0.942      0.605      0.951       0.47
                    b1          5         20      0.996          1      0.995      0.743
                   aew          4         25      0.949          1      0.995      0.784
                   f22          3         17      0.985          1      0.995      0.757
                    p3          6        105          1      0.969      0.995        0.8
                    p8          1          1      0.845          1      0.995      0.796
                   f35          5         32      0.995          1      0.995      0.539
                   f18         13        125      0.984      0.989      0.988      0.827
                   v22          5         41      0.993          1      0.995      0.714
                 su-27          5         31      0.988          1      0.995      0.858
                 il-38         10         27      0.987          1      0.995      0.848
                tu-134          1          1      0.842          1      0.995      0.895
                 su-33          1          2      0.937          1      0.995       0.73
                 an-70          1          2      0.904          1      0.995      0.895
                 tu-22          8         98      0.998          1      0.995      0.834