Posterior temperature optimized Bayesian models for inverse problems in medical imaging
Medical Image Analysis(2022)
摘要
•We present Posterior Temperature Optimized Bayesian Inverse Model (POTOBIM).•A novel approximate Bayesian approach to inverse problems in medical imaging.•We use Bayesian optimization to tune the posterior temperature and the parameters of the prior distribution•We show its superiority to similar approaches on a variety of inverse medical image problems and modalities.
更多查看译文
关键词
Variational inference,Hallucination,Deep learning
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要