关于更换模型去跑的问题

[daydreamertu]

非常感谢您的工作，在阅读之际，依然有一些疑惑望解答，希望能得到你的回复
1、论文中提到可以更换yolov5的网络去跑，请问更换后具体需要做怎么样的处理，是需要更换PGD_yolo中的如下所示的地方吗？还需要做什么修改吗？期待您的答复。

def _forward_once(self, x, profile=False, visualize=False):
    y, dt = [], []  # outputs
    fuse_times = 0
    keypoint_feature = self.HRNETmodel(x)
    for m in self.model:
        if m.f != -1:  # if not from previous layer
            x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f]  # from earlier layers
        if profile:
            self._profile_one_layer(m, x, dt)
        x=m(x)
        if m.i == 38:
        # if m.i == 23:
            x = self.ICAF1([x,keypoint_feature[0]])
            fuse_times += 1               
        if m.i == 41:
        # if m.i == 26:
            x = self.ICAF2([x,keypoint_feature[1]])
            fuse_times += 1                     
        if m.i == 44:
        # if m.i == 29:
            x = self.ICAF3([x,keypoint_feature[2]])
            fuse_times += 1                       
        y.append(x if m.i in self.save else None)  # save output
    assert fuse_times == 3, 'keypoints feature fuse times less than 3'
    return x