Reducing Textural Bias Improves Robustness of Deep Segmentation Models

Autor: Daniel Rueckert, Ahmed E. Fetit, Seoin Chai

Publikováno v: Medical Image Understanding and Analysis ISBN: 9783030804312
MIUA
Annual Conference on Medical Image Understanding and Analysis (MIUA 2021)

Despite advances in deep learning, robustness under domain shift remains a major bottleneck in medical imaging settings. Findings on natural images suggest that deep neural models can show a strong textural bias when carrying out image classification

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::1e91e5fa798f50c7aab41b7ad81d8532