arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2211.00799cs.CVcs.LGeess.IV
收藏

基于双深度图像先验的实用相位恢复方法

Practical Phase Retrieval Using Double Deep Image Priors

University of Minnesota(明尼苏达大学);University of Oxford(牛津大学);Flatiron Institute(熨斗研究院)

Zhong Zhuang, David Yang, Felix Hofmann, David Barmherzig, Ju Sun

摘要:

相位恢复(PR)涉及从复幅度恢复复相位。我们确定PR问题的难度水平和对称性的数量和品种之间的连接。我们专注于最困难的远场PR(FFPR),并提出了一种新的方法,使用双深图像先验。在现实的评估,我们的方法优于所有竞争的方法大幅度。作为一种单实例方法,该方法不需要训练数据和最小的超参数调整,因此具有很好的实用性。

英文摘要:

Phase retrieval (PR) concerns the recovery of complex phases from complex magnitudes. We identify the connection between the difficulty level and the number and variety of symmetries in PR problems. We focus on the most difficult far-field PR (FFPR), and propose a novel method using double deep image priors. In realistic evaluation, our method outperforms all competing methods by large margins. As a single-instance method, our method requires no training data and minimal hyperparameter tuning, and hence enjoys good practicality.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑