search

zanilzanzan / FuseNet_PyTorch

Joint scene classification and semantic segmentation with FuseNet
GNU General Public License v3.0
108 stars 34 forks source link

RuntimeError when first time run the code #2

Closed nessieyang closed 6 years ago

nessieyang commented 6 years ago

I just download the dataset of NYU that you put in the same post. And ran the 'Train_FuseNet.py' on Linux. Then it came to the error like this: image

hazirbas commented 6 years ago

Hi, did you solve the issue? I think it happens once you input a 4-D vector instead of 1-D to depth branch.

Best, Caner

hazirbas commented 6 years ago

Please check the issue #1.

hazirbas commented 6 years ago

Have you solved it? @nessieyang

nessieyang commented 6 years ago

Dear Caner, Thanks very much! I have solved it! The problem occurred at line 36, FuseNetClass.py. The initialization of the first layer have a size mismatch problem. Just use unsqueeze to add a new dimension can solve it.

Best, Yang

hazirbas commented 6 years ago

Hi @nessieyang , can you please create a merge request?