Task-Aware Compressed Sensing with Generative Adversarial Networks

Abstract

In recent years, neural network approaches have been widely adopted for machine learning tasks, with applications in computer vision. More recently, unsupervised generative models based on neural networks have been successfully applied to model data distributions via low-dimensional latent spaces. In this paper, we use Generative Adversarial Networks (GANs) to impose structure in compressed sensing problems, replacing the usual sparsity constraint. We propose to train the GANs in a task-aware fashion, specifically for reconstruction tasks. We also show that it is possible to train our model without using any (or much) non-compressed data. Finally, we show that the latent space of the GAN carries discriminative information and can further be regularized to generate input features for general inference tasks. We demonstrate the effectiveness of our method on a variety of reconstruction and classification problems.

arXiv commentAccepted for publication at the Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18)

Cited by in corpus (12)

• Robust Compressed Sensing MRI with Deep Generative Priors
• Recovering compressed images for automatic crack segmentation using generative models
• Security Matters: A Survey on Adversarial Machine Learning
• Fast Compressive Sensing Recovery Using Generative Models with Structured Latent Variables
• Bayesian Inference with Generative Adversarial Network Priors
• Inference in Multi-Layer Networks with Matrix-Valued Unknowns
• Homotopic Gradients of Generative Density Priors for MR Image Reconstruction
• GAN-based Priors for Quantifying Uncertainty
• Multi-task MR Imaging with Iterative Teacher Forcing and Re-weighted Deep Learning
• Learning Generative Models of Structured Signals from Their Superposition Using GANs with Application to Denoising and Demixing
• Inverse Problems Leveraging Pre-trained Contrastive Representations
• Compressed Sensing via Measurement-Conditional Generative Models