Tensorflow implementation for: "Adversarial Training for Solving Inverse Problems". The code was written by Zhengxia Zou ([email protected]).
Inverse problems are a group of important mathematical problems which aim to estimate the causal factors (e.g., the unknown source data x and parameters z) from a set of observations. Most of the recent approaches consider the inverse problem as a prediction of the source data x based on the observation y with the help of pair-wise training supervision while ignoring the physics behind. We re-examine the inverse problem under a totally different point of view, i.e. instead of predicting x directly from y, we learn to predict the parameters z under an adversarial training paradigm so that to shed light on the causal relationship with a clear physical significance.
With this project, you can train a model to solve the following inverse problems:
- Separating superimposed images on MNIST and CIFAR-10 datasets.
- Image denoising on MNIST
- Removing speckle and streak noise in CAPTCHAs
All the above tasks are trained without any help of pair-wise supervision. For more information of this project, please refer to our paper.
YouTube video: Adversarial Training for Solving Inverse Problems
- python 3.5
- anaconda>=5.2
- opencv>=3.4
- tensorflow >=1.8
Train:
python igan.py
Test:
python deploy.py
Configurations:
vi initializer.py
An illustration of the training process of our method on two groups of 2D data: 1) Gaussian data, and 2) Banana data. Our method aims to learn an invertible mapping (Affine transformation) for each data point so that to map green cluster to the red and inversely map the red to the green. No pair-wise supervision is provided during the training.
Some example results of superimposed image separation by using our method on MNIST and CIFAR-10 datasets. No pair-wise supervision is used during the training.
Some example results of undetermined BBS for CAPTCHA images by using our method. No pair-wise supervision is used during the training.
Some example results of undetermined BSS for speech signals by using our method. Each audio CAPTCHA fragment consists of 4 random pieces of the human voice and is about 2.5 seconds long (8kHz). No pair-wise supervision is used during the training.
Some example results of image denoising on MNIST images. Despite the strong noise (PSNR less than 0.01 dB), our method sill nicely recovers the clear image. No pair-wise supervision is used during the training.
Some example results of image deraining with different methods on the Image Deraining Dataset: CycleGAN, and Pixel-to-Pixel and Ours.
"keep Calm and Don't Overfit."