Coupled VO2 Oscillators Circuit as Analog First Layer Filter in Convolutional Neural Networks

被引:33
作者
Corti, Elisabetta [1 ]
Cornejo Jimenez, Joaquin Antonio [1 ]
Niang, Kham M. [2 ]
Robertson, John [2 ]
Moselund, Kirsten E. [1 ]
Gotsmann, Bernd [1 ]
Ionescu, Adrian M. [3 ]
Karg, Siegfried [1 ]
机构
[1] IBM Res Zurich, Ruschlikon, Switzerland
[2] Univ Cambridge, Dept Engn, Cambridge, England
[3] Ecole Polytech Fed Lausanne, Nanoelect Devices Lab, Lausanne, Switzerland
关键词
oscillatory neural network; vanadium dioxide; phase-encoding; convolutional neural networks; pattern recognition; relaxation oscillators; coupled oscillators;
D O I
10.3389/fnins.2021.628254
中图分类号
Q189 [神经科学];
学科分类号
071006 ;
摘要
In this work we present an in-memory computing platform based on coupled VO2 oscillators fabricated in a crossbar configuration on silicon. Compared to existing platforms, the crossbar configuration promises significant improvements in terms of area density and oscillation frequency. Further, the crossbar devices exhibit low variability and extended reliability, hence, enabling experiments on 4-coupled oscillator. We demonstrate the neuromorphic computing capabilities using the phase relation of the oscillators. As an application, we propose to replace digital filtering operation in a convolutional neural network with oscillating circuits. The concept is tested with a VGG13 architecture on the MNIST dataset, achieving performances of 95% in the recognition task.
引用
收藏
页数:12
相关论文
共 37 条
[1]   Preparation of atomic layer deposited vanadium dioxide thin films using tetrakis(ethylmethylamino) vanadium as precursor [J].
Bai, Guandong ;
Niang, Kham M. ;
Robertson, John .
JOURNAL OF VACUUM SCIENCE & TECHNOLOGY A, 2020, 38 (05)
[2]   Neuromorphic computing with multi-memristive synapses [J].
Boybat, Irem ;
Le Gallo, Manuel ;
Nandakumar, S. R. ;
Moraitis, Timoleon ;
Parnell, Thomas ;
Tuma, Tomas ;
Rajendran, Bipin ;
Leblebici, Yusuf ;
Sebastian, Abu ;
Eleftheriou, Evangelos .
NATURE COMMUNICATIONS, 2018, 9
[3]   Time-Delay Encoded Image Recognition in a Network of Resistively Coupled VO2 on Si Oscillators [J].
Corti, E. ;
Khanna, A. ;
Niang, K. ;
Robertson, J. ;
Moselund, K. E. ;
Gotsmann, B. ;
Datta, S. ;
Karg, S. .
IEEE ELECTRON DEVICE LETTERS, 2020, 41 (04) :629-632
[4]   Scaled resistively-coupled VO2 oscillators for neuromorphic computing [J].
Corti, Elisabetta ;
Gotsmann, Bernd ;
Moselund, Kirsten ;
Ionescu, Adrian M. ;
Robertson, John ;
Karg, Siegfried .
SOLID-STATE ELECTRONICS, 2020, 168
[5]  
Corti E, 2018, 2018 IEEE INTERNATIONAL CONFERENCE ON REBOOTING COMPUTING (ICRC), P195, DOI 10.1109/ICRC.2018.8638626
[6]   Computational Architectures Based on Coupled Oscillators [J].
Cotter, Matthew ;
Fang, Yan ;
Levitan, Steven P. ;
Chiarulli, Donald M. ;
Narayanan, Vijaykrishnan .
2014 IEEE COMPUTER SOCIETY ANNUAL SYMPOSIUM ON VLSI (ISVLSI), 2014, :131-136
[7]   Phase Oscillatory Network and Visual Pattern Recognition [J].
Follmann, Rosangela ;
Macau, Elbert E. N. ;
Rosa, Epaminondas, Jr. ;
Piqueira, Jose R. C. .
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2015, 26 (07) :1539-1544
[8]   Pattern recognition with simple oscillating circuits [J].
Hoelzel, R. W. ;
Krischer, K. .
NEW JOURNAL OF PHYSICS, 2011, 13
[9]   Oscillatory neurocomputers with dynamic connectivity [J].
Hoppensteadt, FC ;
Izhikevich, EM .
PHYSICAL REVIEW LETTERS, 1999, 82 (14) :2983-2986
[10]  
Izhikevich E., 2000, Computing With Oscillators