MT-GAN-PyTorch - PyTorch Implementation of Learning to Transfer: Unsupervised Domain Translation via Meta-Learning

Last update: Oct 19, 2022

Related tags

Deep Learning MT-GAN-PyTorch

Overview

MT-GAN-PyTorch

PyTorch Implementation of AAAI-2020 Paper "Learning to Transfer: Unsupervised Domain Translation via Meta-Learning"

Dependency:

Python 3.6

PyTorch 0.4.0

Usage:

Unsupervised Domain Translation via Meta-Learning

Downloading labels2photos, horses2zebras, summer2winter, apple2orange, monet2photo, cezanne2photo, ukiyoe2photo, vangogh2photo, photos2maps and labels2facades datasets following CycleGAN.

Organize these translation datasets as:

meta_datarooot
├── vangogh2photo
|   ├── trainA
|   ├── trainB
|   
├── ukiyoe2photo
    ├── trainA
    ├── trainB
...

Train MT-GAN on 10-shot tranlation:

$ python train.py --name mtgan_results --model mt_gan --meta_dataroot meta_datarooot --k_spt 10 --k_qry 10 --finetune_step 1000
Test MT-GAN on 10-shot translation:

$ python test.py --name mtgan_results --model mt_gan --meta_dataroot meta_datarooot --k_spt 10 --k_qry 10 --finetune_step 1000
Train MT-GAN on 5-shot translation:

$ python train.py --name mtgan_results --model mt_gan --meta_dataroot meta_datarooot --k_spt 5 --k_qry 5 --finetune_step 1000
Test MT-GAN on 5-shot translation:

$ python test.py --name mtgan_results --model mt_gan --meta_dataroot meta_datarooot --k_spt 5 --k_qry 5 --finetune_step 1000

You might also like...

Code of Adverse Weather Image Translation with Asymmetric and Uncertainty aware GAN

Adverse Weather Image Translation with Asymmetric and Uncertainty-aware GAN (AU-GAN) Official Tensorflow implementation of Adverse Weather Image Trans

36 Dec 26, 2022

Transfer-Learn is an open-source and well-documented library for Transfer Learning.

Transfer-Learn is an open-source and well-documented library for Transfer Learning. It is based on pure PyTorch with high performance and friendly API. Our code is pythonic, and the design is consistent with torchvision. You can easily develop new algorithms, or readily apply existing algorithms.

2.2k Jan 3, 2023

Official Implementation of Domain-Aware Universal Style Transfer

Domain Aware Universal Style Transfer Official Pytorch Implementation of 'Domain Aware Universal Style Transfer' (ICCV 2021) Domain Aware Universal St

80 Dec 30, 2022

Official pytorch implementation of "Feature Stylization and Domain-aware Contrastive Loss for Domain Generalization" ACMMM 2021 (Oral)

Feature Stylization and Domain-aware Contrastive Loss for Domain Generalization This is an official implementation of "Feature Stylization and Domain-

22 Sep 22, 2022

Comments

Regarding loading state_dict in finetuning
Firstly, thank you for making the code public, makes it a lot easier for people who want to apply this method to other domains.

Secondly, I'd like to ask if the following is intentional or a copy-paste error:

netG_A.load_state_dict(self.netG_A.state_dict()) netG_B.load_state_dict(self.netG_A.state_dict()) netD_A.load_state_dict(self.netD_A.state_dict()) netD_B.load_state_dict(self.netD_A.state_dict())

MT-GAN-PyTorch/models/mt_gan_model.py line: 270 -273

Shouldn't netG_B and netD_B be loaded with self.netG_B and self.netD_B state_dicts respectively? If I am not mistakenstate_dict doesn't only contain the architecture information but also the network weights. Also, if it's intentional, then I find it a bit confusing that how models performing opposite operations can share the same weights?
opened by omayrkhan 0
Drawing the Figure 1,2 in the paper and table 1 and table 2

Dear Sir , Brilliant Work! ..thank you very much for your efforts. Please how can I plot ( Figures 1,2 in your paper ) Moreover, how can I get the same results you put in the table 1 and table 2 ?? thanks in advance.

opened by BassantTolba1234 0

MT-GAN-PyTorch - PyTorch Implementation of Learning to Transfer: Unsupervised Domain Translation via Meta-Learning

Related tags

Overview

MT-GAN-PyTorch

Dependency:

Usage:

Unsupervised Domain Translation via Meta-Learning

You might also like...

Code of Adverse Weather Image Translation with Asymmetric and Uncertainty aware GAN

Transfer-Learn is an open-source and well-documented library for Transfer Learning.

Official Implementation of Domain-Aware Universal Style Transfer

Official pytorch implementation of "Feature Stylization and Domain-aware Contrastive Loss for Domain Generalization" ACMMM 2021 (Oral)

A Pytorch Implementation of [Source data‐free domain adaptation of object detector through domain

[CVPR 2021] Unsupervised 3D Shape Completion through GAN Inversion

Code for Dual Contrastive Learning for Unsupervised Image-to-Image Translation, NTIRE, CVPRW 2021.

Implementation of "JOKR: Joint Keypoint Representation for Unsupervised Cross-Domain Motion Retargeting"

Official implementation of "DSP: Dual Soft-Paste for Unsupervised Domain Adaptive Semantic Segmentation"

Comments

Regarding loading state_dict in finetuning

Drawing the Figure 1,2 in the paper and table 1 and table 2

Owner

Implementation of "Meta-rPPG: Remote Heart Rate Estimation Using a Transductive Meta-Learner"

Code for CVPR2021 "Visualizing Adapted Knowledge in Domain Transfer". Visualization for domain adaptation. #explainable-ai

CDTrans: Cross-domain Transformer for Unsupervised Domain Adaptation

CDTrans: Cross-domain Transformer for Unsupervised Domain Adaptation

Pytorch implementation of CVPR2021 paper "MUST-GAN: Multi-level Statistics Transfer for Self-driven Person Image Generation"

Official PyTorch implementation of Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval.

FuseDream: Training-Free Text-to-Image Generationwith Improved CLIP+GAN Space OptimizationFuseDream: Training-Free Text-to-Image Generationwith Improved CLIP+GAN Space Optimization

DR-GAN: Automatic Radial Distortion Rectification Using Conditional GAN in Real-Time

PyTorch Implementation of CycleGAN and SSGAN for Domain Transfer (Minimal)

A PyTorch implementation for Unsupervised Domain Adaptation by Backpropagation(DANN), support Office-31 and Office-Home dataset