Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

UVCGAN: UNet Vision Transformer cycle-consistent GAN for unpaired image-to-image translation

Conference ·

Unpaired image-to-image translation has broad applications in art, design, and scientific simulations. One early breakthrough was CycleGAN that emphasizes one-to-one mappings between two unpaired image domains via generative-adversarial networks (GAN) coupled with the cycle-consistency constraint, while more recent works promote one-to-many mapping to boost diversity of the translated images. Motivated by scientific simulation and one-to-one needs, this work revisits the classic CycleGAN framework and boosts its performance to outperform more contemporary models without relaxing the cycle-consistency constraint. To achieve this, we equip the generator with a Vision Transformer (ViT) and employ necessary training and regularization techniques. Compared to previous best-performing models, our model performs better and retains a strong correlation between the original and translated image. An accompanying ablation study shows that both the gradient penalty and self-supervised pre-training are crucial to the improvement. To promote reproducibility and open science, the source code, hyperparameter configurations, and pre-trained model are available at https: //github.com/LS4GAN/uvcgan.

Research Organization:
Brookhaven National Laboratory (BNL), Upton, NY (United States)
Sponsoring Organization:
Laboratory-Directed Research and Development (LDRD)
DOE Contract Number:
SC0012704
OSTI ID:
1895074
Report Number(s):
BNL-223609-2022-COPA
Country of Publication:
United States
Language:
English

References (24)

The frontier of simulation-based inference journal May 2020
ImageNet: A large-scale hierarchical image database
  • Deng, Jia; Dong, Wei; Socher, Richard
  • 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPR Workshops), 2009 IEEE Conference on Computer Vision and Pattern Recognition https://doi.org/10.1109/CVPR.2009.5206848
conference June 2009
Generative models for molecular discovery: Recent advances and challenges journal March 2022
Countering Malicious DeepFakes: Survey, Battleground, and Horizon journal May 2022
Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks conference October 2017
Image Generators with Conditionally-Independent Pixel Synthesis conference June 2021
Analyzing and Improving the Image Quality of StyleGAN conference June 2020
Context Encoders: Feature Learning by Inpainting conference June 2016
The Cityscapes Dataset for Semantic Urban Scene Understanding conference June 2016
StarGAN v2: Diverse Image Synthesis for Multiple Domains conference June 2020
Transferring GANs: Generating Images from Limited Data book January 2018
iMap: a novel method for statistical fixation mapping of eye movement data journal April 2011
The Creation and Detection of Deepfakes journal January 2021
Multi-task Self-Supervised Visual Learning conference October 2017
Least Squares Generative Adversarial Networks conference October 2017
Deep Learning Face Attributes in the Wild conference December 2015
Breaking the Cycle – Colleagues Are All You Need conference June 2020
Emerging Properties in Self-Supervised Vision Transformers conference October 2021
The SYNTHIA Dataset: A Large Collection of Synthetic Images for Semantic Segmentation of Urban Scenes conference June 2016
MineGAN: Effective Knowledge Transfer From GANs to Target Domains With Few Images conference June 2020
Exploring Simple Siamese Representation Learning conference June 2021
On the Effectiveness of Least Squares Generative Adversarial Networks journal December 2019
Image Generation From Small Datasets via Batch Statistics Adaptation conference October 2019
Unpaired Image-to-Image Translation via Latent Energy Transport conference June 2021

Similar Records

Unpaired image translation to mitigate domain shift in liquid argon time projection chamber detector responses
Journal Article · 2024 · Machine Learning: Science and Technology · OSTI ID:2478421

Hyperparameter Studies for Vision Transformers Trained on High-Fidelity Simulations
Software · 2024 · OSTI ID:code-134596

Potential Flow Generator With L2 Optimal Transport Regularity for Generative Models
Journal Article · 2020 · IEEE Transactions on Neural Networks and Learning Systems · OSTI ID:2281642

Related Subjects