# Cycle-free CycleGAN using Invertible Generator for Unsupervised Low-Dose CT Denoising

arXiv: https://arxiv.org/abs/2104.08538 · HTML: https://mlsanity.com/paper/2104.08538/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 35 references that are in the map.
- First posted on arXiv 17 Apr 2021.

## Builds on (most cited first, 23 of 23)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [Auto-Encoding Variational Bayes](https://mlsanity.com/paper/1312.6114/) (2013)
- [Variational Inference with Normalizing Flows](https://mlsanity.com/paper/1505.05770/) (2015)
- [Density estimation using Real NVP](https://mlsanity.com/paper/1605.08803/) (2016)
- [Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization](https://mlsanity.com/paper/1703.06868/) (2017)
- [Spectral Normalization for Generative Adversarial Networks](https://mlsanity.com/paper/1802.05957/) (2018)
- [Glow: Generative Flow with Invertible 1x1 Convolutions](https://mlsanity.com/paper/1807.03039/) (2018)
- [NICE: Non-linear Independent Components Estimation](https://mlsanity.com/paper/1410.8516/) (2014)
- [Least Squares Generative Adversarial Networks](https://mlsanity.com/paper/1611.04076/) (2016)
- [Computational Optimal Transport](https://mlsanity.com/paper/1803.00567/) (2018)
- [Low-Dose CT with a Residual Encoder-Decoder Convolutional Neural Network (RED-CNN)](https://mlsanity.com/paper/1702.00288/) (2017)
- [Low Dose CT Image Denoising Using a Generative Adversarial Network with Wasserstein Distance and Perceptual Loss](https://mlsanity.com/paper/1708.00961/) (2017)
- [A deep convolutional neural network using directional wavelets for low-dose X-ray CT reconstruction](https://mlsanity.com/paper/1610.09736/) (2016)
- [CT Super-resolution GAN Constrained by the Identical, Residual, and Cycle Learning Ensemble(GAN-CIRCLE)](https://mlsanity.com/paper/1808.04256/) (2018)
- [Understanding and Mitigating Exploding Inverses in Invertible Neural Networks](https://mlsanity.com/paper/2006.09347/) (2020)
- [Unpaired Deep Learning for Accelerated MRI using Optimal Transport Driven CycleGAN](https://mlsanity.com/paper/2008.12967/) (2020)
- [CycleGAN with a Blur Kernel for Deconvolution Microscopy: Optimal Transport Geometry](https://mlsanity.com/paper/1908.09414/) (2019)
- [f-VAEs: Improve VAEs with Conditional Flows](https://mlsanity.com/paper/1809.05861/) (2018)
- [Two-Stage Deep Learning for Accelerated 3D Time-of-Flight MRA without Matched Training Data](https://mlsanity.com/paper/2008.01362/) (2020)
- [General Invertible Transformations for Flow-based Generative Modeling](https://mlsanity.com/paper/2011.15056/) (2020)
- [Unsupervised CT Metal Artifact Learning using Attention-guided beta-CycleGAN](https://mlsanity.com/paper/2007.03480/) (2020)

## Built on by (most cited first, 15 of 15)

- [MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusion](https://mlsanity.com/paper/2203.12621/) (2022)
- [LIT-Former: Linking In-plane and Through-plane Transformers for Simultaneous CT Image Denoising and Deblurring](https://mlsanity.com/paper/2302.10630/) (2023)
- [Synthetic Latent Fingerprint Generator](https://mlsanity.com/paper/2208.13811/) (2022)
- [WIA-LD2ND: Wavelet-based Image Alignment for Self-supervised Low-Dose CT Denoising](https://mlsanity.com/paper/2403.11672/) (2024)
- [Self Supervised Low Dose Computed Tomography Image Denoising Using Invertible Network Exploiting Inter Slice Congruence](https://mlsanity.com/paper/2211.01618/) (2022)
- [Patch-wise Deep Metric Learning for Unsupervised Low-Dose CT Denoising](https://mlsanity.com/paper/2207.02377/) (2022)
- [Multi-Scale Texture Loss for CT denoising with GANs](https://mlsanity.com/paper/2403.16640/) (2024)
- [CM-Diff: A Single Generative Network for Bidirectional Cross-Modality Translation Diffusion Model Between Infrared and Visible Images](https://mlsanity.com/paper/2503.09514/) (2025)
- [Whole-body PET image denoising for reduced acquisition time](https://mlsanity.com/paper/2303.16085/) (2023)
- [GCN-MIF: Graph Convolutional Network with Multi-Information Fusion for Low-dose CT Denoising](https://mlsanity.com/paper/2105.07146/) (2021)
- [Generalizable Denoising of Microscopy Images using Generative Adversarial Networks and Contrastive Learning](https://mlsanity.com/paper/2303.15214/) (2023)
- [A signal processing interpretation of noise-reduction convolutional neural networks](https://mlsanity.com/paper/2307.13425/) (2023)
- [Unsupervised Low-dose CT Reconstruction with One-way Conditional Normalizing Flows](https://mlsanity.com/paper/2410.17543/) (2024)
- [Optimal Transport Driven Asymmetric Image-to-Image Translation for Nuclei Segmentation of Histological Images](https://mlsanity.com/paper/2506.07023/) (2025)
- [One CT Unified Model Training Framework to Rule All Scanning Protocols](https://mlsanity.com/paper/2603.15025/) (2026)

## Abstract (arXiv)

Recently, CycleGAN was shown to provide high-performance, ultra-fast denoising for low-dose X-ray computed tomography (CT) without the need for a paired training dataset. Although this was possible thanks to cycle consistency, CycleGAN requires two generators and two discriminators to enforce cycle consistency, demanding significant GPU resources and technical skills for training. A recent proposal of tunable CycleGAN with Adaptive Instance Normalization (AdaIN) alleviates the problem in part by using a single generator. However, two discriminators and an additional AdaIN code generator are still required for training. To solve this problem, here we present a novel cycle-free Cycle-GAN architecture, which consists of a single generator and a discriminator but still guarantees cycle consistency. The main innovation comes from the observation that the use of an invertible generator automatically fulfills the cycle consistency condition and eliminates the additional discriminator in the CycleGAN formulation. To make the invertible generator more effective, our network is implemented in the wavelet residual domain. Extensive experiments using various levels of low-dose CT images confirm that our method can significantly improve denoising performance using only 10% of learnable parameters and faster training time compared to the conventional CycleGAN.