search

sanghyun-son / EDSR-PyTorch

PyTorch version of the paper 'Enhanced Deep Residual Networks for Single Image Super-Resolution' (CVPRW 2017)
MIT License
2.44k stars 672 forks source link

Change input channel #153

Closed BeepBump closed 5 years ago

BeepBump commented 5 years ago

Hi, Sorry for bothering! If I conduct some transform to the input data, how can I change the number of input channel from 3 to 12? And I need to conduct inverse transform, so the 12-channel output can turn back to 3-channel, where can I modify the code? Pls~Thanks!

sanghyun-son commented 5 years ago

Hello.

You should change the model architecture. I recommend you to create another file for your new model and do experiments with a new model file. Specifically,

  1. Copy src/model/edsr.py to src/model/mymodel.py.

  2. Modify these two lines. Also, class name should be replaced to MyModel instead of EDSR.

  3. Use --model mymodel argument to use your new model.

Thank you!