search

Tramac / awesome-semantic-segmentation-pytorch

Semantic Segmentation on PyTorch (include FCN, PSPNet, Deeplabv3, Deeplabv3+, DANet, DenseASPP, BiSeNet, EncNet, DUNet, ICNet, ENet, OCNet, CCNet, PSANet, CGNet, ESPNet, LEDNet, DFANet)
Apache License 2.0
2.8k stars 581 forks source link

Have the pretrained model been used in eval.py？ #163

Closed mxy519 closed 3 years ago

mxy519 commented 3 years ago

When I trained the model, the validation result is acceptable, but I run the eval.py, the results is too bad, I want to ask: have the pretrained model been used in eval.py？ I do not read the code about pretrained model.

Tramac commented 3 years ago

The pretrained model will be used if pretrained=True.

mxy519 commented 3 years ago

When I trained the model, the validation result is acceptable, but I run the eval.py, the results is too bad, why is it?

Tramac commented 3 years ago

Please confirm whether the pretrained model is loaded normally.

mxy519 commented 3 years ago

How to confirm it? The pretrained is true

mxy519 commented 3 years ago

Thanks, this problem has been solved. Can the backbone be changed to mobilenet?

Tramac commented 3 years ago

Yes, you can refer to this repo, the directory structure is similar.