Orthogonality of Diffractive Deep Neural Networks

Several laws are found for the Diffractive Deep Neural Networks (D2NN). They reveal the inner product of any two light fields in D2NN is invariant and the D2NN act as a unitary transformation for optical fields. If the output intensities of the two inputs are separated spatially, the input fields must be orthogonal. These laws imply that the D2NN is not only suitable for the classification of general objects but also more suitable for applications aim to the optical orthogonal modes. Additionally, our simulation shows D2NN do well in applications like mode conversion, mode multiplexer, and optical mode recognition.

[1]  L Pesquera,et al.  Photonic information processing beyond Turing: an optoelectronic implementation of reservoir computing. , 2012, Optics express.

[2]  D Psaltis,et al.  Optical information processing based on an associative-memory model of neural nets with thresholding and feedback. , 1985, Optics letters.

[3]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[4]  Bram van Ginneken,et al.  A survey on deep learning in medical image analysis , 2017, Medical Image Anal..

[5]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[6]  A. Hirose,et al.  Coherent optical associative memory system that processes complex-amplitude information , 2000, IEEE Photonics Technology Letters.

[7]  Xiaoou Tang,et al.  Image Super-Resolution Using Deep Convolutional Networks , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Xiaogang Wang,et al.  Deep Learning Face Representation from Predicting 10,000 Classes , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[9]  V. G. Krasilenko,et al.  The structures of optical neural nets based on new matrix-tensor equivalental models (MTEMs) and results of modeling , 2010, Optical Memory and Neural Networks.

[10]  Jürgen Schmidhuber,et al.  Multi-column deep neural network for traffic sign classification , 2012, Neural Networks.

[11]  Huug de Waardt,et al.  All fiber-optic neural network using coupled SOA based ring lasers , 2002, IEEE Trans. Neural Networks.

[12]  D Psaltis,et al.  Optical implementation of the Hopfield model. , 1985, Applied optics.

[13]  E C Mos,et al.  Optical-mode neural network by use of the nonlinear response of a laser diode to external optical feedback. , 1997, Applied optics.

[14]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.

[15]  Dirk Englund,et al.  Deep learning with coherent nanophotonic circuits , 2017, 2017 Fifth Berkeley Symposium on Energy Efficient Electronic Systems & Steep Transistors Workshop (E3S).

[16]  A A Friesem,et al.  All-optical neural network with inhibitory neurons. , 1989, Optics letters.

[17]  S. Yeh,et al.  Optical implementation of the hopfield neural network with matrix gratings. , 2004, Applied optics.

[18]  D. Psaltis,et al.  Holography in artificial neural networks , 1990, Nature.