研究文章
一个实时对象探测器基于YOLOv4自驾车辆
|
| 检测算法 |
汽车AP70 |
行人AP50 |
骑自行车AP50 |
地图(%) |
帧/秒 |
输入的大小 |
|
| MS-CNN [14] |
87.42 |
80.43 |
86.28 |
84.71 |
8.13 |
1920×576 |
| SINet [15] |
89.82 |
79.20 |
87.23 |
85.42 |
23.98 |
1920×576 |
| SSD (12] |
85.12 |
48.06 |
50.68 |
61.28 |
28.93 |
512×512 |
| RefineDet [17] |
92.74 |
78.45 |
81.90 |
84.36 |
27.81 |
512×512 |
| CFENet [16] |
88.47 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
512×512 |
| RFBNet [18] |
86.39 |
61.62 |
72.31 |
73.44 |
39.20 |
512×512 |
| YOLOv3 [23] |
79.49 |
79.01 |
83.07 |
80.52 |
43.57 |
512×512 |
| 高斯YOLOv3 [39] |
87.33 |
79.90 |
83.60 |
83.61 |
43.13 |
512×512 |
| YOLOv4 [24] |
90.50 |
80.10 |
88.70 |
86.43 |
52.14 |
512×512 |
| 提出工作(1) |
92.38 |
83.60 |
89.50 |
88.49 |
48.37 |
512×512 |
| 提出工作(2) |
90.05 |
81.10 |
87.90 |
86.35 |
58.47 |
512×512 |
|
|
