Extending Probabilistic U-Net Using MC-Dropout to Quantify Data and Model Uncertainty

BRAINLESION: GLIOMA, MULTIPLE SCLEROSIS, STROKE AND TRAUMATIC BRAIN INJURIES, BRAINLES 2021, PT II（2022）

引用 0|浏览3

暂无评分

摘要

We extend the Probabilistic U-Net using MC-Dropout to estimate model uncertainty in addition to the data uncertainty in order to improve the overall predictive uncertainty estimate. We use this model on the datasets present in the QUBIQ21 challenge and achieve a mean score of 0.719.

查看译文

关键词

model uncertainty,quantify data,u-net,mc-dropout

AI 理解论文

溯源树

样例

生成溯源树，研究论文发展脉络

Chat Paper

正在生成论文摘要