Phase imaging with an untrained neural network
Chinese Academy of Sciences · Shanghai Institute of Optics and Fine Mechanics · +4 more institutions
Abstract
Most of the neural networks proposed so far for computational imaging (CI) in optics employ a supervised training strategy, and thus need a large training set to optimize their weights and biases. Setting aside the requirements of environmental and system stability during many hours of data acquisition, in many practical applications, it is unlikely to be possible to obtain sufficient numbers of ground-truth images for training. Here, we propose to overcome this limitation by incorporating into a conventional deep neural network a complete physical model that represents the process of image formation. The most significant advantage of the resulting physics-enhanced deep neural network (PhysenNet) is that it…
Citation impact
- FWCI
- 54.50
- Percentile
- 100%
- References
- 36
Authors
8- FWFei WangCorresponding
Chinese Academy of Sciences, Shanghai Institute of Optics and Fine Mechanics, University of Chinese Academy of Sciences
- YBYaoming Bian
Chinese Academy of Sciences, Shanghai Institute of Optics and Fine Mechanics, University of Chinese Academy of Sciences
- HWHaichao Wang
Chinese Academy of Sciences, Shanghai Institute of Optics and Fine Mechanics, University of Chinese Academy of Sciences
- MLMeng Lyu
Shanghai Institute of Optics and Fine Mechanics, University of Chinese Academy of Sciences
- GPGiancarlo Pedrini
University of Stuttgart
Topics & keywords
- Artificial neural network
- Computer science
- Artificial intelligence
- Process (computing)
- Object (grammar)
- Set (abstract data type)
- Time delay neural network
- Stability (learning theory)
- Life in Land