反向传播
人工神经网络
衍射
物理
深层神经网络
瑞利散射
计算机科学
人工智能
光学
作者
Kohei Yamamoto,Hiroyuki Yanagisawa
摘要
We have applied physics aware training (PAT) to diffractive deep neural networks (D2NN) consisting of multiple spatial light modulators (SLMs) to close the reality gap between the simulation model and the physical system. Compared to conventional training methods using only simulation models, PAT improves classification accuracy in the experiment. In this method, an analytic expression for backpropagation is based on Rayleigh-Sommerfeld diffraction integral as conventional, but the backpropagated error values are replaced by the measured values.
科研通智能强力驱动
Strongly Powered by AbleSci AI