AttributeError: 'Conv1dBlock' object has no attribute 'conv_fn' about upidet

[Eaphan]

Sorry for the mistake, you can pull the latest change and run it again.

[Z-Lee-corder]

Hello, I have rerun the program according to your update. The following error still occurs:

Traceback (most recent call last):
File "/media/lizheng/Samsung/codes/BiProDet/tools/train.py", line 202, in
main()
File "/media/lizheng/Samsung/codes/BiProDet/tools/train.py", line 154, in main
train_model(
File "/media/lizheng/Samsung/codes/BiProDet/tools/train_utils/train_utils.py", line 111, in train_model
accumulated_iter = train_one_epoch(
File "/media/lizheng/Samsung/codes/BiProDet/tools/train_utils/train_utils.py", line 47, in train_one_epoch
loss, tb_dict, disp_dict = model_func(model, batch)
File "/media/lizheng/Samsung/codes/BiProDet/tools/../pcdet/models/init.py", line 42, in model_func
ret_dict, tb_dict, disp_dict = model(batch_dict)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/media/lizheng/Samsung/codes/BiProDet/tools/../pcdet/models/detectors/point_3dssd.py", line 17, in forward
batch_dict = cur_module(batch_dict)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/media/lizheng/Samsung/codes/BiProDet/tools/../pcdet/models/backbones_3d/pointnet2_backbone.py", line 541, in forward
li_xyz, li_features, li_vis, li_scores, image_x = self.SA_modules[i](
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/media/lizheng/Samsung/codes/BiProDet/tools/../pcdet/ops/pointnet2/pointnet2_batch/pointnet2_modules.py", line 923, in forward
feat_2d_to_3d = self.fuse3d_before_mlps(feat_2d_to_3d)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/container.py", line 141, in forward
input = module(input)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/media/lizheng/Samsung/codes/BiProDet/tools/../pcdet/ops/pointnet2/pointnet2_batch/pointnet2_utils.py", line 575, in forward
x = self.norm(x)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/batchnorm.py", line 135, in forward
self._check_input_dim(input)
File "/home/lizheng/anaconda3/envs/open-mmlab/lib/python3.8/site-packages/torch/nn/modules/batchnorm.py", line 408, in _check_input_dim
raise ValueError("expected 4D input (got {}D input)".format(input.dim()))
ValueError: expected 4D input (got 3D input)

[Eaphan]

You can pull the latest change and run it again.

In a commit, we wrongly define the Conv1dBlock function.