Photonic Long-Short Term Memory Neural Networks with Analog Memory

A Photonic implementation is proposed for the Long-Short Term Memory neural network, offering fundamental speed and bandwidth advantages over digital electronic implementations. Integrated analog memory for photonics is designed as a component of this network.

[1]  Florent Altché,et al.  An LSTM network for highway trajectory prediction , 2017, 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC).

[2]  Jürgen Schmidhuber,et al.  LSTM: A Search Space Odyssey , 2015, IEEE Transactions on Neural Networks and Learning Systems.

[3]  Paul R. Prucnal,et al.  Broadcast and Weight: An Integrated Network For Scalable Photonic Spike Processing , 2014, Journal of Lightwave Technology.

[4]  Ellen Zhou,et al.  Neuromorphic photonic networks using silicon photonic weight banks , 2017, Scientific Reports.

[5]  Silvio Savarese,et al.  Social LSTM: Human Trajectory Prediction in Crowded Spaces , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[6]  Volker J. Sorger,et al.  Electronic Bottleneck Suppression in Next-generation Networks with Integrated Photonic Digital-to-analog Converters , 2019 .

[7]  Paul R. Prucnal,et al.  Progress in neuromorphic photonics , 2017 .

[8]  Paul R. Prucnal,et al.  Noise Analysis of Photonic Modulator Neurons , 2019, IEEE Journal of Selected Topics in Quantum Electronics.

[9]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[10]  Nikos Pleros,et al.  Optical RAM and integrated optical memories: a survey , 2020, Light: Science & Applications.

[11]  Lukas Chrostowski,et al.  Silicon Photonics Circuit Design: Methods, Tools and Challenges , 2018 .