https://github.com/Megvii-BaseDetection/BEVDepth/blob/d78c7b58b10b9ada940462ba83ab24d99cae5833/bevdepth/layers/backbones/fusion_lss_fpn.py#L62 这里两个都是[0,0]有问题吗? aug_scale = torch.sqrt(mats_dict['ida_mats'][:, 0, :, 0, 0]**2 + mats_dict['ida_mats'][:, 0, :, 0, 0]**2).reshape(-1, 1)