Error information 1
experiments/mode.py#171
code:
ref_node_corr_indices, src_node_corr_indices, node_corr_scores = self.coarse_matching(
ref_feats_c_norm, src_feats_c_norm, ref_node_masks, src_node_masks, mask
)error:
Traceback (most recent call last):
File "trainval.py", line 66, in <module>
main()
File "trainval.py", line 62, in main
trainer.run()
File "/data16t/xxx/HybridReg_PyTorch/geotransformer/engine/epoch_based_trainer.py", line 180, in run
self.train_epoch()
File "/data16t/xxx/HybridReg_PyTorch/geotransformer/engine/epoch_based_trainer.py", line 95, in train_epoch
output_dict, result_dict = self.train_step(self.epoch, self.inner_iteration, data_dict)
File "trainval.py", line 43, in train_step
output_dict = self.model(data_dict)
File "/data/xxx/miniconda3/envs/hybridreg/lib/python3.8/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/data16t/xxx/HybridReg_PyTorch/experiments/hybridmatch/model.py", line 176, in forward
ref_node_corr_indices, src_node_corr_indices, node_corr_scores = self.coarse_matching(
File "/data/xxx/miniconda3/envs/hybridreg/lib/python3.8/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
TypeError: forward() takes from 3 to 5 positional arguments but 6 were given
The mask parameter is redundant in the SuperPointMatching module. The parameters of 'SuperPointMatching.forward()' are:
class SuperPointMatching(nn.Module):
def __init__(self, num_correspondences, dual_normalization=True):
super(SuperPointMatching, self).__init__()
self.num_correspondences = num_correspondences
self.dual_normalization = dual_normalization
def forward(self, ref_feats, src_feats, ref_masks=None, src_masks=None):
r"""Extract superpoint correspondences.
Args:
ref_feats (Tensor): features of the superpoints in reference point cloud.
src_feats (Tensor): features of the superpoints in source point cloud.
ref_masks (BoolTensor=None): masks of the superpoints in reference point cloud (False if empty).
src_masks (BoolTensor=None): masks of the superpoints in source point cloud (False if empty).
Returns:
ref_corr_indices (LongTensor): indices of the corresponding superpoints in reference point cloud.
src_corr_indices (LongTensor): indices of the corresponding superpoints in source point cloud.
corr_scores (Tensor): scores of the correspondences.
"""
if ref_masks is None:
ref_masks = torch.ones(size=(ref_feats.shape[0],), dtype=torch.bool).cuda()
if src_masks is None:
src_masks = torch.ones(size=(src_feats.shape[0],), dtype=torch.bool).cuda()
...Error information 2
experiments/hybridmatch/loss.py#92
code:
aplace_loss, corr_loss, mask_loss = self.laplace_loss(output_dict, data_dict)
The returned variables self.laplace_loss(LaplaceLoss) are
...
if self.stage == 1:
corr_sp_mask = torch.ones(corr_gt.shape[0] + corr_gt.shape[1], device=device)
else:
corr_sp_mask = output_dict['corr_sp_mask']
loss = self.loss(corr_gt, corr_es, (1 - corr_sp_mask))
return loss, 1-gt_maskThe variables are not matched between the defined version and the captured version.
summary
Maybe the version of geotransformer module is not matched with the code in hybridmatch experiment.
Sincere thanks to the author for reading and answering my questions in time, and correcting possible errors.
Error information 1
experiments/mode.py#171code:
error:
The
maskparameter is redundant in theSuperPointMatchingmodule. The parameters of 'SuperPointMatching.forward()' are:Error information 2
experiments/hybridmatch/loss.py#92code:
The returned variables
self.laplace_loss(LaplaceLoss)are... if self.stage == 1: corr_sp_mask = torch.ones(corr_gt.shape[0] + corr_gt.shape[1], device=device) else: corr_sp_mask = output_dict['corr_sp_mask'] loss = self.loss(corr_gt, corr_es, (1 - corr_sp_mask)) return loss, 1-gt_maskThe variables are not matched between the defined version and the captured version.
summary
Maybe the version of
geotransformermodule is not matched with the code inhybridmatchexperiment.Sincere thanks to the author for reading and answering my questions in time, and correcting possible errors.