Nittscher, M; Lameter, M; Barbano, R; Leuschner, J; Jin, B; Maass, P; (2023) SVD-DIP: Overcoming the Overfitting Problem in DIP-based CT Reconstruction. In: Oguz, Ipek and Noble, Jack and Li, Xiaoxiao and Styner, Martin and Baumgartner, Christian and Rusu, Mirabela and Heinmann, Tobias and Kontos, Despinas and Landman, Bennett and Dawant, Benoit, (eds.) Proceedings of Machine Learning Research. (pp. pp. 617-642). ML Research Press: Nashville, TN, USA.

Preview

Text nittscher24a.pdf - Published Version Download (4MB) | Preview

Abstract

The deep image prior (DIP) is a well-established unsupervised deep learning method for image reconstruction; yet it is far from being flawless. The DIP overfits to noise if not early stopped, or optimized via a regularized objective. We build on the regularized fine-tuning of a pretrained DIP, by adopting a novel strategy that restricts the learning to the adaptation of singular values. The proposed SVD-DIP uses ad hoc convolutional layers whose pretrained parameters are decomposed via the singular value decomposition. Optimizing the DIP then solely consists in the fine-tuning of the singular values, while keeping the left and right singular vectors fixed. We thoroughly validate the proposed method on real-measured µCT data of a lotus root as well as two medical datasets (LoDoPaB and Mayo). We report significantly improved stability of the DIP optimization, by overcoming the overfitting to noise.

Download activity - last month

Export as XMLJSONCSV

Download activity - last 12 months

Export as CSVJSONXML

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item