DOCSLIB.ORG

Silicon Photonics for Artificial Intelligence Applications

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text

Load more

FOCUS PHOTONICS AND ARTIFICIAL INTELLIGENCE SILICON PHOTONICS FOR ARTIFICIAL INTELLIGENCE APPLICATIONS /////////////////////////////////////////////////////////////////////////////////////////////////// Bicky A. MARQUEZ1 ,* , Matthew J. FILIPOVICH1, Emma R. HOWARD1, Viraj BANGARI1, Zhimu GUO1, Hugh D. MORISON1, Thomas FERREIRA DE LIMA2, Alexander N. TAIT2,3, Paul R. PRUCNAL2, Bhavin J. SHASTRI 1,2 1 Department of Physics, Engineering Physics & Astronomy, Queen’s University, Kingston, ON KL7 3N6, Canada 2 Department of Electrical Engineering, Princeton University, Princeton, NJ 08544, USA 3 Applied Physics Division, National Institute of Standards and Technology, Boulder, CO 80305, USA *[email protected] Artificial intelligence enabled by neural networks has enabled applications in many fields (e.g. medicine, finance, autonomous vehicles). Software implementations of neural networks on conventional computers are limited in speed and energy efficiency. Neuromorphic engineering aims to build processors in which hardware mimic neurons and synapses in brain for distributed and parallel processing. Neuromorphic engineering enabled by silicon photonics can offer sub- nanosecond latencies, and can extend the domain of artificial intelligence applications to high-performance computing and ultrafast learning. We discuss current progress and challenges on these demonstrations to scale to practical systems for training and inference. https://doi.org/10.1051/photon/202010440 nalog computing common neural models are spiking The photonic platform is currently has recently been artificial neurons and perceptrons. one of the most promising technolo- considered a poten- While spiking artificial neurons are gies to tackle the expensive calcula- tial avenue to de- biologically realistic, the field of AI is tions performed by deep networks. crease energy and currently perceptron-based [1]. Silicon photonics offers high-scala- time requirements Perceptrons implement multi- bility, high-bandwidth, low-footprint, for executing algorithms such as ply-accumulate (MAC) operations. and low-energy consumption [3]. The deepA neural networks. Analog spe- MAC operations serve to quantify the high-bandwidth and multiwavelen- cial-purpose hardware requires number of multiplications and addi- gth parallel properties of light allow the manufacturing of machines to tions required to run deep networks. for optical information processing at physically model each individual A perceptron of M inputs can per- a high data rate. The ability of neuro- component of such networks. This form M MAC operations per time morphic photonic systems to provide proves to be a significant challenge step. Multiple MAC operations can substantial improvement in our com- as current deep networks scale up be executed in parallel to implement puting capabilities is moving ever to thousands or even billions of any type of artificial neural network closer, with, potentially, PetaMac/ neurons to solve complex artificial (ANN). MACs are currently the most second/mm2 processing speeds. intelligence (AI) tasks. To enable the burdensome hardware bottleneck in In this article, we describe a photo- use of analog machines to map brain ANNs; for instance, the deep network nic scheme that can perform parallel circuitry, the functions of biological AlexNet requires 724 million MACs to MAC operations on-chip and introduce neurons must be modelled. The most solve ImageNet [2]. two photonic platforms that allow for 40 www.photoniques.com I Photoniques 104 PHOTONICS AND ARTIFICIAL INTELLIGENCE FOCUS AI hardware acceleration: i) a special-pur- approximately 100 billion neurons and pose photonic architecture for executing 100 trillion synaptic interconnections the direct feedback alignment (DFA) al- that must be represented in an artificial gorithm for neural network training [6], machine. However, a subset of the brain and ii) an implementation of a Long-Short circuitry can still be represented in an Term Memory (LSTM) neural network [7]. artificial machine to simulate some of Both proposed designs offer fundamental the human cognitive processes. speed and bandwidth advantages over di- Recently, most significant advances in gital electronic implementations. the field of AI have been achieved using a perceptron, shown in Fig. 1, as the ar- BACKGROUND: NEUROSCIENCE tificial model of the neuron. The output y AND COMPUTATION of the neuron represents the signals sent Digital computers are typically computing from the axon of a biological neuron and systems that perform logical and mathe- is mathematically described by matical operations with high accuracy. y = f (W∙ x+b). Nowadays, such complex systems signi- ficantly outweigh human capabilities for The xi inputs transmit the information calculation and memory. Nevertheless, to the neuron through the weights Wi if we were to compare a human agent which correspond to the strength of the with a digital machine, we would see that synapses. The summation of all weighted there are many abstractions that should inputs and their transformation via acti- be made to perform one-to-one compa- vation function f are associated with the risons. Such abstractions assume that physiological role of the neuron’s cell human cognitive processes are comple- body. The bias b represents an extra va- tely procedural and follow standard logic. riable that remains in the system even if However, most human cognitive acts do the other inputs are absent. not follow a set of well-defined instruc- ANNs are built using perceptrons as tions. Therefore, a one-to-one mapping neural primitives such that the synaptic between human and digital computers connections are either positive or nega- might not be suitable. tive to mimic excitatory and inhibitory Analog neuromorphic computing ap- neural behaviour. A nonlinear activation proaches might be more suited to mimic function can be used to define activated human brain processes. The goal is to and deactivated behaviours in artificial create a one-to-one mapping between the neurons. ANNs can be categorized as neural system and the analog machine, either feed forward (where connections where each biological quantity is mo- between neurons do not form a cycle) delled by an equivalent analog artificial or recurrent neural networks (where model. For an architecture such as the cycles exist). human brain, this could be a demanding Attempts to build fast and efficient requirement. The human brain contains perceptron-based ANNs have been Figure 1. Schematic diagram of a perceptron. Photoniques 104 I www.photoniques.com 41 FOCUS Photonics and artificial intelligence reported throughout recent years. An Figure 2. An array of M MRRs can emulate interesting computing acceleration Add-drop MRR weight bank with a balanced the weighted addition of a single neu- technique consists in using hardware photodetector implementing M element-wise ron if add-drop MRRs and a balanced units to perform MAC operations at multipliers to perform N MAC operations photodetector are incorporated into high speeds. A MAC unit performs in parallel. the model, as shown in Fig. 2. In this multiplications and accumulation illustration we show how to perform processes: (a+w.x). Multiple MAC M MAC operations in parallel in pho- operations can be run in parallel to electro-optically, or through light tonics. Input values to the neuron perform complex operations such as absorption such as phase-change or can be mapped to voltage values Vi convolutions and digital filters. MACs graphene materials. In this work, tu- that tune each individual MRR(xi). are typically used in implementations ning is performed by thermally mo- Each voltage value has a one-to-one of ANNs in digital electronics [4]. difying the refractive index of the MRR correspondence with an MRR trans- Nevertheless, the serialization of waveguide. The application of voltage mission profile Ti, and the same the summands to perform weighted values to the heater allows us to map principle holds for weight values. The addition makes this process ineffi- real-valued numbers to the device. experimental implementation of this cient; consequently, chip designers method requires the use of M lasers are looking for alternative solutions with different wavelengthsλ i (with such as full parallelism. One of the Figure 3. i = 1, … ,M) that represent M channels. most promising technologies for (a) Transmission versus wavelength Two MRRs with different on and this purpose is based on the photo- curves of two different MRRs (MRR(x1), off resonance configurations at the nic platform. MRR(W1) performing element-wise optical same wavelength λ1 will therefore per- multiplications, and (b) the product of form element-wise multiplications, PHOTONIC PERCEPTRON such multiplication. as shown in Fig. 3. Here, we show AND MAC OPERATIONS an illustration of the multiplication A scalable photonic architecture that implements parallel MACs can be achieved using on-chip wavelength division multiplexing (WDM) tech- niques [8]. This design uses microring resonators (MRRs) [9], i.e. photonic synapses, to encode input values and weights onto multiple wavelength si- gnals. Tuning a given MRR on and off resonance changes the transmission of each signal through the respec- tive filter, effectively multiplying the signal with a desired weight. An ad- vantage of using MRRs is the ability to tune the weight values using a va- riety of different methods: thermally, 42 www.photoniques.com I Photoniques 104 PHOTONICS AND ARTIFICIAL INTELLIGENCE FOCUS between two transmission
Recommended publications
  • Neural Network Transformation and Co-Design Under Neuromorphic Hardware Constraints

    Neural Network Transformation and Co-Design Under Neuromorphic Hardware Constraints

    NEUTRAMS: Neural Network Transformation and Co-design under Neuromorphic Hardware Constraints Yu Ji ∗, YouHui Zhang∗‡, ShuangChen Li†,PingChi†, CiHang Jiang∗,PengQu∗,YuanXie†, WenGuang Chen∗‡ Email: [email protected], [email protected] ∗Department of Computer Science and Technology, Tsinghua University, PR.China † Department of Electrical and Computer Engineering, University of California at Santa Barbara, USA ‡ Center for Brain-Inspired Computing Research, Tsinghua University, PR.China Abstract—With the recent reincarnations of neuromorphic (CMP)1 with a Network-on-Chip (NoC) has emerged as a computing comes the promise of a new computing paradigm, promising platform for neural network simulation, motivating with a focus on the design and fabrication of neuromorphic chips. many recent studies of neuromorphic chips [4]–[10]. The A key challenge in design, however, is that programming such chips is difficult. This paper proposes a systematic methodology design of neuromorphic chips is promising to enable a new with a set of tools to address this challenge. The proposed computing paradigm. toolset is called NEUTRAMS (Neural network Transformation, One major issue is that neuromorphic hardware usually Mapping and Simulation), and includes three key components: places constraints on NN models (such as the maximum fan- a neural network (NN) transformation algorithm, a configurable in/fan-out of a single neuron, the limited range of synaptic clock-driven simulator of neuromorphic chips and an optimized runtime tool that maps NNs onto the target hardware for weights) that it can support, and the hardware types of neurons better resource utilization. To address the challenges of hardware or activation functions are usually simpler than the software constraints on implementing NN models (such as the maximum counterparts.
  • Deep Learning with Tensorflow 2 and Keras Second Edition

    Deep Learning with Tensorflow 2 and Keras Second Edition

    Deep Learning with TensorFlow 2 and Keras Second Edition Regression, ConvNets, GANs, RNNs, NLP, and more with TensorFlow 2 and the Keras API Antonio Gulli Amita Kapoor Sujit Pal BIRMINGHAM - MUMBAI Deep Learning with TensorFlow 2 and Keras Second Edition Copyright © 2019 Packt Publishing All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without the prior written permission of the publisher, except in the case of brief quotations embedded in critical articles or reviews. Every effort has been made in the preparation of this book to ensure the accuracy of the information presented. However, the information contained in this book is sold without warranty, either express or implied. Neither the authors, nor Packt Publishing or its dealers and distributors, will be held liable for any damages caused or alleged to have been caused directly or indirectly by this book. Packt Publishing has endeavored to provide trademark information about all of the companies and products mentioned in this book by the appropriate use of capitals. However, Packt Publishing cannot guarantee the accuracy of this information. Commissioning Editor: Amey Varangaonkar Acquisition Editors: Yogesh Deokar, Ben Renow-Clarke Acquisition Editor – Peer Reviews: Suresh Jain Content Development Editor: Ian Hough Technical Editor: Gaurav Gavas Project Editor: Janice Gonsalves Proofreader: Safs Editing Indexer: Rekha Nair Presentation Designer: Sandip Tadge First published: April 2017 Second edition: December 2019 Production reference: 2130320 Published by Packt Publishing Ltd. Livery Place 35 Livery Street Birmingham B3 2PB, UK. ISBN 978-1-83882-341-2 www.packt.com packt.com Subscribe to our online digital library for full access to over 7,000 books and videos, as well as industry leading tools to help you plan your personal development and advance your career.
  • A Sparsity-Driven Backpropagation-Less Learning Framework Using Populations of Spiking Growth Transform Neurons

    A Sparsity-Driven Backpropagation-Less Learning Framework Using Populations of Spiking Growth Transform Neurons

    ORIGINAL RESEARCH published: 28 July 2021 doi: 10.3389/fnins.2021.715451 A Sparsity-Driven Backpropagation-Less Learning Framework Using Populations of Spiking Growth Transform Neurons Ahana Gangopadhyay and Shantanu Chakrabartty* Department of Electrical and Systems Engineering, Washington University in St. Louis, St. Louis, MO, United States Growth-transform (GT) neurons and their population models allow for independent control over the spiking statistics and the transient population dynamics while optimizing a physically plausible distributed energy functional involving continuous-valued neural variables. In this paper we describe a backpropagation-less learning approach to train a Edited by: network of spiking GT neurons by enforcing sparsity constraints on the overall network Siddharth Joshi, spiking activity. The key features of the model and the proposed learning framework University of Notre Dame, United States are: (a) spike responses are generated as a result of constraint violation and hence Reviewed by: can be viewed as Lagrangian parameters; (b) the optimal parameters for a given task Wenrui Zhang, can be learned using neurally relevant local learning rules and in an online manner; University of California, Santa Barbara, United States (c) the network optimizes itself to encode the solution with as few spikes as possible Youhui Zhang, (sparsity); (d) the network optimizes itself to operate at a solution with the maximum Tsinghua University, China dynamic range and away from saturation; and (e) the framework is flexible enough to *Correspondence: incorporate additional structural and connectivity constraints on the network. As a result, Shantanu Chakrabartty [email protected] the proposed formulation is attractive for designing neuromorphic tinyML systems that are constrained in energy, resources, and network structure.
  • Multilayer Spiking Neural Network for Audio Samples Classification Using

    Multilayer Spiking Neural Network for Audio Samples Classification Using

    Multilayer Spiking Neural Network for Audio Samples Classification Using SpiNNaker Juan Pedro Dominguez-Morales, Angel Jimenez-Fernandez, Antonio Rios-Navarro, Elena Cerezuela-Escudero, Daniel Gutierrez-Galan, Manuel J. Dominguez-Morales, and Gabriel Jimenez-Moreno Robotic and Technology of Computers Lab, Department of Architecture and Technology of Computers, University of Seville, Seville, Spain {jpdominguez,ajimenez,arios,ecerezuela, dgutierrez,mdominguez,gaji}@atc.us.es http://www.atc.us.es Abstract. Audio classification has always been an interesting subject of research inside the neuromorphic engineering field. Tools like Nengo or Brian, and hard‐ ware platforms like the SpiNNaker board are rapidly increasing in popularity in the neuromorphic community due to the ease of modelling spiking neural networks with them. In this manuscript a multilayer spiking neural network for audio samples classification using SpiNNaker is presented. The network consists of different leaky integrate-and-fire neuron layers. The connections between them are trained using novel firing rate based algorithms and tested using sets of pure tones with frequencies that range from 130.813 to 1396.91 Hz. The hit rate percentage values are obtained after adding a random noise signal to the original pure tone signal. The results show very good classification results (above 85 % hit rate) for each class when the Signal-to-noise ratio is above 3 decibels, vali‐ dating the robustness of the network configuration and the training step. Keywords: SpiNNaker · Spiking neural network · Audio samples classification · Spikes · Neuromorphic auditory sensor · Address-Event Representation 1 Introduction Neuromorphic engineering is a discipline that studies, designs and implements hardware and software with the aim of mimicking the way in which nervous systems work, focusing its main inspiration on how the brain solves complex problems easily.
  • Neuromorphic Computing Using Emerging Synaptic Devices: a Retrospective Summary and an Outlook

    Neuromorphic Computing Using Emerging Synaptic Devices: a Retrospective Summary and an Outlook

    electronics Review Neuromorphic Computing Using Emerging Synaptic Devices: A Retrospective Summary and an Outlook Jaeyoung Park 1,2 1 School of Computer Science and Electronic Engineering, Handong Global University, Pohang 37554, Korea; [email protected]; Tel.: +82-54-260-1394 2 School of Electronic Engineering, Soongsil University, Seoul 06978, Korea Received: 29 June 2020; Accepted: 5 August 2020; Published: 1 September 2020 Abstract: In this paper, emerging memory devices are investigated for a promising synaptic device of neuromorphic computing. Because the neuromorphic computing hardware requires high memory density, fast speed, and low power as well as a unique characteristic that simulates the function of learning by imitating the process of the human brain, memristor devices are considered as a promising candidate because of their desirable characteristic. Among them, Phase-change RAM (PRAM) Resistive RAM (ReRAM), Magnetic RAM (MRAM), and Atomic Switch Network (ASN) are selected to review. Even if the memristor devices show such characteristics, the inherent error by their physical properties needs to be resolved. This paper suggests adopting an approximate computing approach to deal with the error without degrading the advantages of emerging memory devices. Keywords: neuromorphic computing; memristor; PRAM; ReRAM; MRAM; ASN; synaptic device 1. Introduction Artificial Intelligence (AI), called machine intelligence, has been researched over the decades, and the AI now becomes an essential part of the industry [1–3]. Artificial intelligence means machines that imitate the cognitive functions of humans such as learning and problem solving [1,3]. Similarly, neuromorphic computing has been researched that imitates the process of the human brain compute and store [4–6].

  • Neuromorphic Processing and Sensing

    1 Neuromorphic Processing & Sensing: Evolutionary Progression of AI to Spiking Philippe Reiter, Geet Rose Jose, Spyridon Bizmpikis, Ionela-Ancuța Cîrjilă machine learning and deep learning (MLDL) developments. Abstract— The increasing rise in machine learning and deep This accelerated pace of innovation was spurred on by the learning applications is requiring ever more computational seminal works of LeCun, Hinton and others in the 1990s on resources to successfully meet the growing demands of an always- convolutional neural networks, or CNNs. Since then, numerous, connected, automated world. Neuromorphic technologies based on more advanced learning models have been developed, and Spiking Neural Network algorithms hold the promise to implement advanced artificial intelligence using a fraction of the neural networks have become integral to industry, medicine, computations and power requirements by modeling the academia and commercial devices. From fully autonomous functioning, and spiking, of the human brain. With the vehicles to rapid facial recognition entering popular culture to proliferation of tools and platforms aiding data scientists and enumerable innovations across almost all domains, it is not an machine learning engineers to develop the latest innovations in exaggeration to claim that CNNs and their progeny have artificial and deep neural networks, a transition to a new paradigm will require building from the current well-established changed both the technological and physical worlds. foundations. This paper explains the theoretical workings of CNNs can be exceedingly computationally intensive, neuromorphic technologies based on spikes, and overviews the however, as will be explained, often limiting their use to high- state-of-art in hardware processors, software platforms and performance computers and large data centres.
  • Neuromorphic Computing in Ginzburg-Landau Lattice Systems

    Neuromorphic Computing in Ginzburg-Landau Lattice Systems

    Neuromorphic computing in Ginzburg-Landau polariton lattice systems Andrzej Opala and Michał Matuszewski Institute of Physics, Polish Academy of Sciences, Warsaw Sanjib Ghosh and Timothy C. H. Liew Division of Physics and Applied Physics, Nanyang Technological University, Singapore < Andrzej Opala and Michał Matuszewski Institute of Physics, Polish Academy of Sciences, Warsaw Sanjib Ghosh and Timothy C. H. Liew Division of Physics and Applied Physics, Nanyang Technological University, Singapore < Machine learning with neural networks What is machine learning? Modern computers are extremely efficient in solving many problems, but some tasks are very hard to implement. ● It is difficult to write an algorithm that would recognize an object from different viewpoints or in di erent sceneries, lighting conditions etc. ● It is difficult to detect a fradulent credit card transaction. !e need to combine a large number of weak rules with complex dependencies rather than follow simple and reliable rules. "hese tasks are often relatively easy for humans. Machine learning with neural networks Machine learning algorithm is not written to solve a speci#c task, but to learn how to classify / detect / predict on its own. ● A large number of examples is collected and used as an input for the algorithm during the teaching phase. ● "he algorithm produces a program &eg. a neural network with #ne tuned weights), which is able to do the job not only on the provided samples, but also on data that it has never seen before. It usually contains a great number of parameters. Machine learning with neural networks "hree main types of machine learning: 1.Supervised learning – direct feedback, input data is provided with clearly de#ned labels.
  • Spinnaker Project

    Spinnaker Project

    10/21/19 The SpiNNaker project Steve Furber ICL Professor of Computer Engineering The University of Manchester 1 200 years ago… • Ada Lovelace, b. 10 Dec. 1815 "I have my hopes, and very distinct ones too, of one day getting cerebral phenomena such that I can put them into mathematical equations--in short, a law or laws for the mutual actions of the molecules of brain. …. I hope to bequeath to the generations a calculus of the nervous system.” 2 1 10/21/19 70 years ago… 3 Bio-inspiration • Can massively-parallel computing resources accelerate our understanding of brain function? • Can our growing understanding of brain function point the way to more efficient parallel, fault-tolerant computation? 4 2 10/21/19 ConvNets - structure • Dense convolution kernels • Abstract neurons • Only feed-forward connections • Trained through backpropagation 5 The cortex - structure Feedback input Feed-forward output Feed-forward input Feedback output • Spiking neurons • Two-dimensional structure • Sparse connectivity 6 3 10/21/19 ConvNets - GPUs • Dense matrix multiplications • 3.2kW • Low precision 7 Cortical models - Supercomputers • Sparse matrix operations • Efficient communication of spikes • 2.3MW 8 4 10/21/19 Cortical models - Neuromorphic hardware • Memory local to computation • Low-power • Real time • 62mW 9 Start-ups and industry interest 5 10/21/19 SpiNNaker project • A million mobile phone processors in one computer • Able to model about 1% of the human brain… • …or 10 mice! 11 SpiNNaker system 12 6 10/21/19 SpiNNaker chip Multi-chip packaging
  • 1. Introduction Recently, a New Discipline Has Come up Which

    1. Introduction Recently, a New Discipline Has Come up Which

    JOURNAL OF REAL-TIME IMAGE PROCESSING SPECIAL ISSUE ON REAL-TIME NEUROMORPHIC SYSTEMS FOR IMAGE PROCESSING 1. Introduction Recently, a new discipline has come up which challenges classical approaches to computer vision research and engineering, the so-called neuromorphic engineering. This new research paradigm develops engineering systems with unparalleled robustness and on-line adaptability. Such systems are based on algorithms and techniques closely replicating and inspired from the principles of information processing in biological nervous systems. The robustness of the human visual system in almost any visual situation is enviable, performing enormous calculation tasks continuously, robustly, and effortlessly. Their applicability to real-time computer vision challenges is extended, and includes prosthesis systems, robust human-computer interfaces, smart sensors, implanted electronic devices, autonomous visually-guided robotics systems. Neuromorphic engineering fills the gap between, on the one hand, computational neuroscience, and, on the other hand, traditional engineering. Computational neuroscience has yielded very useful models and theories of brain function, although often these have been restricted to simplified conditions, stimuli and tasks, to allow direct comparison with simple empirical measurements on biological systems. Also, computational neuroscience models typically are concerned with testing one given hypothesis, and thus are not intended to solve real-world problems, but they advance the understanding of brain function
  • Embodied Neuromorphic Vision with Event-Driven Random Backpropagation

    Embodied Neuromorphic Vision with Event-Driven Random Backpropagation

    1 Embodied Neuromorphic Vision with Event-Driven Random Backpropagation Jacques Kaiser1, Alexander Friedrich1, J. Camilo Vasquez Tieck1, Daniel Reichard1, Arne Roennau1, Emre Neftci2, Rüdiger Dillmann1 Abstract—Spike-based communication between biological neu- rons is sparse and unreliable. This enables the brain to process visual information from the eyes efficiently. Taking inspiration from biology, artificial spiking neural networks coupled with silicon retinas attempt to model these computations. Recent findings in machine learning allowed the derivation of a family of powerful synaptic plasticity rules approximating backpropaga- tion for spiking networks. Are these rules capable of processing real-world visual sensory data? In this paper, we evaluate the performance of Event-Driven Random Back-Propagation (eRBP) at learning representations from event streams provided by a Dynamic Vision Sensor (DVS). First, we show that eRBP matches state-of-the-art performance on the DvsGesture dataset with the addition of a simple covert attention mechanism. By remapping visual receptive fields relatively to the center of the motion, this attention mechanism provides translation invariance at low computational cost compared to convolutions. Second, we successfully integrate eRBP in a real robotic setup, where a robotic arm grasps objects according to detected visual affordances. In this setup, visual information is actively sensed Figure 1: Our robotic setup embodying the synaptic learning by a DVS mounted on a robotic head performing microsaccadic rule eRBP [1]. The DVS is mounted on a robotic head per- eye movements. We show that our method classifies affordances forming microsaccadic eye movements. The spiking network within 100ms after microsaccade onset, which is comparable to is trained (dashed-line connections) in a supervised fashion human performance reported in behavioral study.
  • CV’20 , ECCV Demo, Arxiv’21 ]

    CV’20 , ECCV Demo, Arxiv’21 ]

    Chankyu Lee Phone: (+1) 765-337-4551 Email: [email protected] Website: https://chan8972.github.io WORK SUMMARY Broadly speaking, my research interests lie at the intersection of deep learning and edge computing. I focus on developing energy-efficient and robust deep learning algorithms with special interests in Spiking Neural Networks (SNNs) and computer vision for event-based cameras. Specifically, I designed unsupervised/supervised/semi-supervised/self-supervised learning algorithms for deep SNNs. In addi- tion, I developed motion estimation algorithms for event-based camera in challenging scenes such as high speed and wide dynamic range. RESEARCH FOCUSES Machine/Deep Learning Computer Vision Algorithm-Hardware Co-design EDUCATION Purdue University, West Lafayette, IN, USA Fall 2015 - Spring 2021 Ph.D., Electrical and Computer Engineering (Advisor: Prof. Kaushik Roy) Sungkyunkwan University (SKKU), South Korea Spring 2009 - Spring 2015 B.S., Electrical and Electronics Engineering Hong Kong University of Science and Technology (HKUST), Hong Kong Fall 2013 Exchange Student Program, Electronic and Computer Engineering EXPERIENCE Center for Brain Inspired Computing (C-BRIC), Purdue University Aug. 2016 - Dec. 2020 Graduate Research Assistant West Lafayette, IN, USA • Exploratory research on neuromorphic computing for energy-efficient and robust machine learning, over- coming limitations of current artificial intelligence through algorithm-hardware co-design. { Proposed novel unsupervised/supervised/semi-supervised learning algorithms for deep convolutional Spiking Neural Networks (SNNs) to efficiently harness machine learning algorithms in the resource constrained real-world environments. Results are presented in publications [IEEE TCDS, Frontiers'18 , Frontiers'20 , arXiv'20 ]. { Developed energy-efficient motion estimation algorithms for event-based camera in challenging scenes such as high speed and wide dynamic range.
  • A Hybrid CMOS-Memristor Neuromorphic Synapse Mostafa Rahimi Azghadi, Member, IEEE, Bernabe Linares-Barranco, Fellow, IEEE, Derek Abbott, Fellow, IEEE, Philip H.W

    A Hybrid CMOS-Memristor Neuromorphic Synapse Mostafa Rahimi Azghadi, Member, IEEE, Bernabe Linares-Barranco, Fellow, IEEE, Derek Abbott, Fellow, IEEE, Philip H.W

    IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS 1 A Hybrid CMOS-memristor Neuromorphic Synapse Mostafa Rahimi Azghadi, Member, IEEE, Bernabe Linares-Barranco, Fellow, IEEE, Derek Abbott, Fellow, IEEE, Philip H.W. Leong, Senior Member, IEEE Abstract—Although data processing technology continues to of synapses and their role in large-scale learning, there is advance at an astonishing rate, computers with brain-like pro- still a need to implement a versatile memristive synapse cessing capabilities still elude us. It is envisioned that such that is capable of faithfully reproducing a larger regime computers may be achieved by the fusion of neuroscience and nano-electronics to realize a brain-inspired platform. This paper of experimental data that takes into account conventional proposes a high-performance nano-scale Complementary Metal STDP [12], frequency-dependent STDP [13], triplet [14], [15] Oxide Semiconductor (CMOS)-memristive circuit, which mimics and quadruplet [15], [16] plasticity experiments. In a recent a number of essential learning properties of biological synapses. study, Wei et al. replicated the outcome of a variety of synaptic The proposed synaptic circuit that is composed of memristors and plasticity experiments including STDP, frequency-dependent CMOS transistors, alters its memristance in response to timing differences among its pre- and post-synaptic action potentials, STDP, triplet, and quadruplet spike interactions, using a TiO2 giving rise to a family of Spike Timing Dependent Plasticity memristor [17]. (STDP). The presented design advances preceding memristive This paper proposes a new hybrid CMOS-memristive circuit synapse designs with regards to the ability to replicate essential that aims to emulate all the aforementioned experimental data, behaviours characterised in a number of electrophysiological with minimal errors close to those reported in a phenomeno- experiments performed in the animal brain, which involve higher order spike interactions.