YOLOv3 can detect objects according to three scales as shown in Figure 1(a). Its feature extraction network comes from Darknet-53, which adopts the full.
確定! 回上一頁