skip to main content
Language:
Search Limited to: Search Limited to: Resource type Show Results with: Show Results with: Search type Index

On instabilities of deep learning in image reconstruction and the potential costs of AI

Proceedings of the National Academy of Sciences - PNAS, 2020-12, Vol.117 (48), p.30088-30095 [Peer Reviewed Journal]

Copyright National Academy of Sciences Dec 1, 2020 ;info:eu-repo/semantics/openAccess ;2020 ;ISSN: 0027-8424 ;EISSN: 1091-6490 ;DOI: 10.1073/pnas.1907377117 ;PMID: 32393633

Full text available

Citations Cited by
  • Title:
    On instabilities of deep learning in image reconstruction and the potential costs of AI
  • Author: Antun, Vegard ; Renna, Francesco ; Poon, Clarice ; Adcock, Ben ; Hansen, Anders C.
  • Subjects: Algorithms ; COLLOQUIUM PAPERS ; Deep learning ; Government agencies ; Image classification ; Image processing ; Image reconstruction ; Machine learning ; on the Science of Deep Learning ; Physical Sciences ; Stability tests
  • Is Part Of: Proceedings of the National Academy of Sciences - PNAS, 2020-12, Vol.117 (48), p.30088-30095
  • Description: Deep learning, due to its unprecedented success in tasks such as image classification, has emerged as a new tool in image reconstruction with potential to change the field. In this paper, we demonstrate a crucial phenomenon: Deep learning typically yields unstable methods for image reconstruction. The instabilities usually occur in several forms: 1) Certain tiny, almost undetectable perturbations, both in the image and sampling domain, may result in severe artefacts in the reconstruction; 2) a small structural change, for example, a tumor, may not be captured in the reconstructed image; and 3) (a counterintuitive type of instability) more samples may yield poorer performance. Our stability test with algorithms and easy-to-use software detects the instability phenomena. The test is aimed at researchers, to test their networks for instabilities, and for government agencies, such as the Food and Drug Administration (FDA), to secure safe use of deep learning methods.
  • Publisher: United States: National Academy of Sciences
  • Language: English;Norwegian
  • Identifier: ISSN: 0027-8424
    EISSN: 1091-6490
    DOI: 10.1073/pnas.1907377117
    PMID: 32393633
  • Source: GFMER Free Medical Journals
    NORA Norwegian Open Research Archives
    PubMed Central
Back to results list
INSPIRE LIBRARY - TON DUC THANG UNIVERSITY (84-028) 37 755 057 Feedback
19 Nguyen Huu Tho St. Dist.7, HCM thuvien@tdtu.edu.vn

Copyright © 2017 INSPIRE LIBRARY - TON DUC THANG UNIVERSITY

Searching Remote Databases, Please Wait