Učitavanje...

An Analysis of the Vulnerability of Two Common Deep Learning-Based Medical Image Segmentation Techniques to Model Inversion Attacks

Recent research in computer vision has shown that original images used for training of deep learning models can be reconstructed using so-called inversion attacks. However, the feasibility of this attack type has not been investigated for complex 3D medical images. Thus, the aim of this study was to...

Cijeli opis

Spremljeno u:
Bibliografski detalji
Izdano u:Sensors (Basel)
Glavni autori: Subbanna, Nagesh, Wilms, Matthias, Tuladhar, Anup, Forkert, Nils D.
Format: Artigo
Jezik:Inglês
Izdano: MDPI 2021
Teme:
Online pristup:https://ncbi.nlm.nih.gov/pmc/articles/PMC8199987/
https://ncbi.nlm.nih.gov/pubmed/34199735
https://ncbi.nlm.nih.govhttp://dx.doi.org/10.3390/s21113874
Oznake: Dodaj oznaku
Bez oznaka, Budi prvi tko označuje ovaj zapis!