The Realization Of Parallel Subtraction Using Lclv In Optical Neural Network

Bin Wang,Jianwen Yang,Xu Liu,Peifu Gu,Jinfa Tang,Yueguang Zhang
DOI: https://doi.org/10.1117/12.252943
1996-01-01
Abstract:Basing on the method of wavelength and beam direction encoding for implementing a bipolar optical neural network, a liquid crystal light value (LCLV) is used to encode the positive and negative light spots by adding relevant light biases so as to obtain the normal and inverse states, respectively. Then, the two modulated light spots in the same group are moved together by using a grating or a lenslet array. The overlapped light spot whose power is just the expected subtractive results of positive and negative weighted summation can be obtained. Taking advantage of said method, parallel operation between positive and negative light spots can be fulfilled. Due to the high discrimination and quick responsibility of LCLV, system operation speed can be greatly enhanced in optical neural network on a large scale. The feasibility is demonstrated by the preliminary experiment.
What problem does this paper attempt to address?