- Improved GANLoss(using Feature Maching Loss(Multi-Scale Discriminator Loss))
- VGGLoss
- l1Loss or Smoothl1Loss
- Wandb monitoring.
- Pix2PixHD Network & Loss.
- Dataloader Using class __ getitem __.
- Efficient parameter parsing to source code.
- Environment : Windows10, 3090RTX, pytorch-nightly
- Trained_model_link : https://www.dropbox.com/s/1bomldev67zhee0/tested.zip?dl=0
- Predicted_Submission.npy_link : https://www.dropbox.com/s/psp507e2d4q1ce7/21110389_submit.npy?dl=0
This repo was referenced by https://github.com/sjmin99/Ther2RGB-Translation
InClass : https://github.com/sejongresearch/ComputerVision
EvalAIStarters : https://github.com/sejong-rcv/EvalAI-Starters