Training Auto-encoder-based Optimizers for Terahertz Image Reconstruction

  title={Training Auto-encoder-based Optimizers for Terahertz Image Reconstruction},
  author={Tak Ming Wong and Matthias Kahl and Peter Haring Bol{\'i}var and Andreas Kolb and Michael M{\"o}ller},
Terahertz (THz) sensing is a promising imaging technology for a wide variety of different applications. Extracting the interpretable and physically meaningful parameters for such applications, however, requires solving an inverse problem in which a model function determined by these parameters needs to be fitted to the measured data. Since the underlying optimization problem is nonconvex and very costly to solve, we propose learning the prediction of suitable parameters from the measured data… 
2 Citations
High-resolution range profile reconstruction method for terahertz FMCW radar.
A software method is proposed to reconstruct the range profile, yielding significant improvements in range resolution and in signal parameter extraction, and the multiple signal characterization algorithm is introduced to indicate a higher-resolution range profile qualitatively.
Seeing through a Black Box: Toward High-Quality Terahertz TomographicImaging via Multi-Scale Spatio-Spectral Image Fusion
This paper proposes a novel multi-scale spatio-spectral fusion Unet (MS3Unet) that extracts multi- scale features from the different spectral of terahertz image data for restoration and demonstrates the effectiveness of the proposed MS3-Unet image restoration approach on 3-D terAhertz tomographic reconstruction applications.


Learning Proximal Operators: Using Denoising Networks for Regularizing Inverse Imaging Problems
This paper studies the possibility of replacing the proximal operator of the regularization used in many convex energy minimization algorithms by a denoising neural network, and obtains state-of-the-art reconstruction results.
One Network to Solve Them All — Solving Linear Inverse Problems Using Deep Projection Models
This work proposes a general framework to train a single deep neural network that solves arbitrary linear inverse problems and demonstrates superior performance over traditional methods using wavelet sparsity prior while achieving performance comparable to specially-trained networks on tasks including compressive sensing and pixel-wise inpainting.
Controlling Neural Networks via Energy Dissipation
This work proposes energy dissipating networks that iteratively compute a descent direction with respect to a given cost function or energy at the currently estimated reconstruction and proves that under standard assumptions, descent using the direction predicted by the network converges (linearly) to the global minimum of the energy.
MoFA: Model-Based Deep Convolutional Face Autoencoder for Unsupervised Monocular Reconstruction
A novel model-based deep convolutional autoencoder that addresses the highly challenging problem of reconstructing a 3D human face from a single in-the-wild color image and can be trained end-to-end in an unsupervised manner, which renders training on very large real world data feasible.
Deep Decoder: Concise Image Representations from Untrained Non-convolutional Networks
This paper proposes an untrained simple image model, called the deep decoder, which is a deep neural network that can generate natural images from very few weight parameters, with a simple architecture with no convolutions and fewer weight parameters than the output dimensionality.
Deep Image Prior
It is shown that a randomly-initialized neural network can be used as a handcrafted prior with excellent results in standard inverse problems such as denoising, superresolution, and inpainting.
OptNet: Differentiable Optimization as a Layer in Neural Networks
OptNet is presented, a network architecture that integrates optimization problems (here, specifically in the form of quadratic programs) as individual layers in larger end-to-end trainable deep networks, and shows how techniques from sensitivity analysis, bilevel optimization, and implicit differentiation can be used to exactly differentiate through these layers.
Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring
This work proposes a multi-scale convolutional neural network that restores sharp images in an end-to-end manner where blur is caused by various sources and presents a new large-scale dataset that provides pairs of realistic blurry image and the corresponding ground truth sharp image that are obtained by a high-speed camera.
Adam: A Method for Stochastic Optimization
This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.
Terahertz image super-resolution based on a deep convolutional neural network.
Experimental results on real THz images show that the proposed method significantly improves the quality ofTHz images with increased resolution and decreased noise, which proves the practicability and exactitude of the proposedmethod.