Author:
Perry Altai,Weng Xiaojing,Feng Ji,Vuong Luat T.
Abstract
Encoded-diffraction hybrid systems—optical encoding and simple electronic decoding—offers feature distillation via model training. Additionally, the most faithfully reconstructed images are not the ones that are best classified. We parametrize our results with singular value decomposition (SVD) entropy, a proxy for image complexity.
Reference7 articles.
1. Inference in artificial intelligence with deep optics and photonics
2. Image sensing with multilayer nonlinear optical neural networks
3. Razlighi Q. R. and Kehtarnavaz N., “A comparison study of image spatial entropy,” in Visual Communications and Image Processing 2009, Rabbani M. and Stevenson R. L., eds. (SPIE, 2009).
4. Madrid-Herrera L., Chacon-Murguia M. I., Posada-Urrutia D. A., and Ramirez-Quintana J. A., “Human Image Complexity Analysis Using a Fuzzy Inference System,” in 2019 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE, 2019).
5. Perry A., Weng X., Nozari E., and Vuong L., “Parameterized Learning and Distillation with Vortex-encoded Spectral Correlations,” (2023).