☆ 3.8 Proceedings Paper

ANOMALY DETECTION IN EM IMAGES - A ZERO-SHOT LEARNING APPROACH

2022 IEEE INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (IEEE ISBI 2022) (2022)

GENERALIZE ULTRASOUND IMAGE SEGMENTATION VIA INSTANT AND PLUG & PLAY STYLE TRANSFER

Zhendong Liu et al.

Summary: This paper proposed a robust segmentation method for unknown appearance shifts by embedding hierarchical style transfer units and adopting Dynamic Instance Normalization, which can achieve precise and dynamic style transfer in a fast and lightweight manner. Extensive experiments on a large dataset from three vendors demonstrate that the proposed method enhances the robustness of deep segmentation models.

2021 IEEE 18TH INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (ISBI) (2021)

添加到收藏夹

Proceedings Paper Computer Science, Artificial Intelligence

Style Curriculum Learning for Robust Medical Image Segmentation

Zhendong Liu et al.

Summary: This paper proposes a new framework to address the issue of distribution shifts in image intensities between training and test data sets in deep segmentation models. The framework includes a novel curriculum design inspired by curriculum learning, an automated gradient manipulation method, and a Local Gradient Sign strategy. Extensive experiments on the public M&Ms Challenge dataset show that the proposed framework can generalize deep models well to unknown distributions and achieve significant improvements in segmentation accuracy.

MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT I (2021)

添加到收藏夹

Article Engineering, Biomedical