Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Reehorst_thesis.pdf (2.61 MB)

ETD Abstract Container

Abstract Header

Machine Learning for Image Inverse Problems and Novelty Detection

Reehorst, Edward Thomas
http://rave.ohiolink.edu/etdc/view?acc_num=osu1669977051952055

Abstract Details

2022, Doctor of Philosophy, Ohio State University, Electrical and Computer Engineering.
This dissertation addresses two separate engineering challenges: image-inverse problems and novelty detection. First, we address image-inverse problems. We review Plug-and-Play (PnP) algorithms, where a proximal operator is replaced by a call of an arbitrary denoising algorithm. We apply PnP algorithms to compressive Magnetic Resonance Imaging (MRI). MRI is a non-invasive diagnostic tool that provides excellent soft-tissue contrast without the use of ionizing radiation. However, when compared to other clinical imaging modalities (e.g., CT or ultrasound), the data acquisition process for MRI is inherently slow, which motivates undersampling and thus drives the need for ac- curate, efficient reconstruction methods from undersampled datasets. We apply the PnP-ADMM algorithm to cardiac MRI and knee MRI data. For these algorithms, we developed learned denoisers that can process complex-valued MRI images. Our algorithms achieve state-of-the-art performance on both the cardiac and knee datasets. Regularization by Denoising (RED), as proposed by Romano, Elad, and Milanfar, is a powerful image-recovery framework that aims to minimize an explicit regular- ization objective constructed from a plug-in image-denoising function. Experimental evidence suggests that RED algorithms are state-of-the-art. We claim, however, that explicit regularization does not explain the RED algorithms. In particular, we show that many of the expressions in the paper by Romano et al. hold only when the denoiser has a symmetric Jacobian, and we demonstrate that such symmetry does not occur with practical denoisers such as non-local means, BM3D, TNRD, and DnCNN. To explain the RED algorithms, we propose a new framework called Score-Matching by Denoising (SMD), which aims to match a “score” (i.e., the gradient of a log-prior). Novelty detection is the ability for a machine learning system to detect signals that are significantly different from samples seen during training. Detecting novelties is important in any problem where it is possible for the system to encounter data that is significantly different from the training data. In recent years, novelty detection in high dimensional signals has been an area of significant research, with many methods proposed. In order to leverage this tremendous body of work, we develop, Shift- Ensembled Novelty Detection (SEND), a framework for combining multiple novelty detection methods. Our experiments show SEND achieves state-of-the-art performance for novelty detection of images. We also apply SEND to the problem of radio frequency (RF) waveform novelty detection. Recently, there has been tremendous research interest in utilizing self-supervised neural networks for novelty detection. These methods have largely been developed for image applications. In this work, we adapt SimCLR and CSI, two popular self-supervised learning methods, to operate with RF waveform data. Our experiments for RF waveform novelty detection show a significant increase in performance over state-of-the-art methods.
Philip Schniter (Advisor)
Rizwan Ahmad (Committee Member)
Lee Potter (Committee Member)
152 p.
Electrical Engineering
Machine Learning; Image Inverse Problems; Compressive Imaging; Denoising; Novelty Detection

Recommended Citations

Citations

  • Reehorst, E. T. (2022). Machine Learning for Image Inverse Problems and Novelty Detection [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1669977051952055

    APA Style (7th edition)

  • Reehorst, Edward. Machine Learning for Image Inverse Problems and Novelty Detection. 2022. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1669977051952055.

    MLA Style (8th edition)

  • Reehorst, Edward. "Machine Learning for Image Inverse Problems and Novelty Detection." Doctoral dissertation, Ohio State University, 2022. http://rave.ohiolink.edu/etdc/view?acc_num=osu1669977051952055

    Chicago Manual of Style (17th edition)

osu1669977051952055
147
© 2022, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.
Release 3.2.12