Runtimerror

[MrDec]

Traceback (most recent call last):
File "/home/dww/OD/others/SRGCAE-master/train_SRGCAE_Local.py", line 153, in
train_model(args)
File "/home/dww/OD/others/SRGCAE-master/train_SRGCAE_Local.py", line 76, in train_model
feat_t1 = GCAE_model(node_t1, norm_adj_t1)
File "/home/dww/anaconda3/envs/detection/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/home/dww/OD/others/SRGCAE-master/model/SRGCAE.py", line 35, in forward
x = torch.sigmoid(self.gc1(x, adj))
File "/home/dww/anaconda3/envs/detection/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/home/dww/OD/others/SRGCAE-master/model/GraphConv.py", line 33, in forward
output = torch.mm(adj, support)
RuntimeError: mat1 and mat2 shapes cannot be multiplied (358x358 and 0x16)

运行报错

[ChenHongruixuan]

Hi @MrDec ,

This problem should be caused by the different version of slic algorithm. Please modify it to: objects = slic(img_t2, n_segments=args.n_seg, compactness=args.cmp, start_label=0).

In recent version, the label returned by slic starts from 1 instead of 0. The latter is the returned start result of the version in our code environment.

Best,