REMODEL: Rethinking Deep CNN Models to Detect and Count on a NeuroSynaptic System

Shukla, Rohit; Lipasti, Mikko; Van Essen, Brian; Moody, Adam; Maruyama, Naoya

doi:10.3389/fnins.2019.00004

REMODEL: Rethinking Deep CNN Models to Detect and Count on a NeuroSynaptic System

Journal Article · Fri Feb 22 00:00:00 EST 2019 · Frontiers in Neuroscience (Online)

DOI:https://doi.org/10.3389/fnins.2019.00004· OSTI ID:2375413

Shukla, Rohit ^[1]; Lipasti, Mikko ^[1]; Van Essen, Brian ^[2]; Moody, Adam ^[2]; Maruyama, Naoya ^[2]

Univ. of Wisconsin, Madison, WI (United States)
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)

In this work, we perform analysis of detection and counting of cars using a low-power IBM TrueNorth Neurosynaptic System. For our evaluation we looked at a publicly-available dataset that has overhead imagery of cars with context present in the image. The trained neural network for image analysis was deployed on the NS16e system using IBM's EEDN training framework. Through multiple experiments we identify the architectural bottlenecks present in TrueNorth system that does not let us deploy large neural network structures. Following these experiments we propose changes to CNN model to circumvent these architectural bottlenecks. The results of these evaluations have been compared with caffe-based implementations of standard neural networks that were deployed on a Titan-X GPU. Results showed that TrueNorth can detect cars from the dataset with 97.60% accuracy and can be used to accurately count the number of cars in the image with 69.04% accuracy. The car detection accuracy and car count (–/+ 2 error margin) accuracy are comparable to high-precision neural networks like AlexNet, GoogLeNet, and ResCeption, but show a manifold improvement in power consumption.

View Accepted Manuscript (DOE)

Research Organization:: Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)

Sponsoring Organization:: USDOE National Nuclear Security Administration (NNSA)

Grant/Contract Number:: AC52-07NA27344

OSTI ID:: 2375413

Report Number(s):: LLNL--JRNL-767281; 957572

Journal Information:: Frontiers in Neuroscience (Online), Journal Name: Frontiers in Neuroscience (Online) Journal Issue: 4 Vol. 13; ISSN 1662-453X

Publisher:: Frontiers Research FoundationCopyright Statement

Country of Publication:: United States

Language:: English

References (25)

Spiking Deep Convolutional Neural Networks for Energy-Efficient Object Recognition Cao, Yongqiang; Chen, Yang; Khosla, Deepak International Journal of Computer Vision, Vol. 113, Issue 1 https://doi.org/10.1007/s11263-014-0788-3	journal	November 2014
Energy-efficient neural network chips approach human recognition capabilities Maass, Wolfgang Proceedings of the National Academy of Sciences, Vol. 113, Issue 41 https://doi.org/10.1073/pnas.1614109113	journal	October 2016
Wafer-scale integration of analog neural networks Schemmel, Johannes; Fieres, Johannes; Meier, Karlheinz 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence) https://doi.org/10.1109/ijcnn.2008.4633828	conference	June 2008
The SpiNNaker Project Furber, Steve B.; Galluppi, Francesco; Temple, Steve Proceedings of the IEEE, Vol. 102, Issue 5 https://doi.org/10.1109/jproc.2014.2304638	journal	May 2014
Controlled Flight of a Biologically Inspired, Insect-Scale Robot Ma, K. Y.; Chirarattananon, P.; Fuller, S. B. Science, Vol. 340, Issue 6132 https://doi.org/10.1126/science.1231806	journal	May 2013
Converting Static Image Datasets to Spiking Neuromorphic Datasets Using Saccades Orchard, Garrick; Jayawant, Ajinkya; Cohen, Gregory K. Frontiers in Neuroscience, Vol. 9 https://doi.org/10.3389/fnins.2015.00437	journal	November 2015
CIFAR10-DVS: An Event-Stream Dataset for Object Classification Li, Hongmin; Liu, Hanchao; Ji, Xiangyang Frontiers in Neuroscience, Vol. 11 https://doi.org/10.3389/fnins.2017.00309	journal	May 2017
Computing Generalized Matrix Inverse on Spiking Neural Substrate Shukla, Rohit; Khoram, Soroosh; Jorgensen, Erik Frontiers in Neuroscience, Vol. 12 https://doi.org/10.3389/fnins.2018.00115	journal	March 2018
Deep Learning With Spiking Neurons: Opportunities and Challenges Pfeiffer, Michael; Pfeil, Thomas Frontiers in Neuroscience, Vol. 12 https://doi.org/10.3389/fnins.2018.00774	journal	October 2018
Striving for Simplicity: The All Convolutional Net Springenberg, Jost Tobias; Dosovitskiy, Alexey; Brox, Thomas arXiv https://doi.org/10.48550/arxiv.1412.6806	preprint	January 2014
A Large Contextual Dataset for Classification, Detection and Counting of Cars with Deep Learning Mundhenk, T. Nathan; Konjevod, Goran; Sakla, Wesam A. arXiv https://doi.org/10.48550/arxiv.1609.04453	preprint	January 2016
Discovering Low-Precision Networks Close to Full-Precision Networks for Efficient Embedded Inference McKinstry, Jeffrey L.; Esser, Steven K.; Appuswamy, Rathinakumar arXiv https://doi.org/10.48550/arxiv.1809.04191	preprint	January 2018
Deep learning LeCun, Yann; Bengio, Yoshua; Hinton, Geoffrey Nature, Vol. 521, Issue 7553 https://doi.org/10.1038/nature14539	journal	May 2015
Convolutional networks for fast, energy-efficient neuromorphic computing Esser, Steven K.; Merolla, Paul A.; Arthur, John V. Proceedings of the National Academy of Sciences, Vol. 113, Issue 41 https://doi.org/10.1073/pnas.1604850113	journal	September 2016
Spiking neural network (SNN) control of a flapping insect-scale robot Clawson, Taylor S.; Ferrari, Silvia; Fuller, Sawyer B. 2016 IEEE 55th Conference on Decision and Control (CDC) https://doi.org/10.1109/CDC.2016.7798778	conference	December 2016
Cognitive computing building block: A versatile and efficient digital neuron model for neurosynaptic cores Cassidy, Andrew S.; Merolla, Paul; Arthur, John V. 2013 International Joint Conference on Neural Networks (IJCNN 2013 - Dallas), The 2013 International Joint Conference on Neural Networks (IJCNN) https://doi.org/10.1109/IJCNN.2013.6707077	conference	August 2013
TrueHappiness: Neuromorphic emotion recognition on TrueNorth Diehl, Peter U.; Pedroni, Bruno U.; Cassidy, Andrew 2016 International Joint Conference on Neural Networks (IJCNN) https://doi.org/10.1109/IJCNN.2016.7727758	conference	July 2016
Evaluating hopfield-network-based linear solvers for hardware constrained neural substrates Shukla, Rohit; Jorgensen, Erik; Lipasti, Mikko 2017 International Joint Conference on Neural Networks (IJCNN) https://doi.org/10.1109/IJCNN.2017.7966352	conference	May 2017
The Dark Side of DNN Pruning Yazdani, Reza; Riera, Marc; Arnau, Jose-Maria 2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture (ISCA) https://doi.org/10.1109/ISCA.2018.00071	conference	June 2018
The SpiNNaker Project Furber, Steve B.; Galluppi, Francesco; Temple, Steve Proceedings of the IEEE, Vol. 102, Issue 5 https://doi.org/10.1109/JPROC.2014.2304638	journal	May 2014
TrueNorth Ecosystem for Brain-Inspired Computing: Scalable Systems, Software, and Applications Sawada, Jun; Akopyan, Filipp; Cassidy, Andrew S. SC16: International Conference for High Performance Computing, Networking, Storage and Analysis https://doi.org/10.1109/SC.2016.11	conference	November 2016
TrueNorth: Design and Tool Flow of a 65 mW 1 Million Neuron Programmable Neurosynaptic Chip Akopyan, Filipp; Sawada, Jun; Cassidy, Andrew IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Vol. 34, Issue 10 https://doi.org/10.1109/TCAD.2015.2474396	journal	October 2015
A million spiking-neuron integrated circuit with a scalable communication network and interface Merolla, P. A.; Arthur, J. V.; Alvarez-Icaza, R. Science, Vol. 345, Issue 6197 https://doi.org/10.1126/science.1254642	journal	August 2014
A new learning method for inference accuracy, core occupation, and performance co-optimization on TrueNorth chip Wen, Wei; Wu, Chunpeng; Wang, Yandan Proceedings of the 53rd Annual Design Automation Conference https://doi.org/10.1145/2897937.2897968	conference	June 2016
Conversion of Continuous-Valued Deep Networks to Efficient Event-Driven Networks for Image Classification Rueckauer, Bodo; Lungu, Iulia-Alexandra; Hu, Yuhuang Frontiers in Neuroscience, Vol. 11 https://doi.org/10.3389/fnins.2017.00682	journal	December 2017

Cited By (1)

Convolutional-Neural Network-Based Image Crowd Counting: Review, Categorization, Analysis, and Performance Evaluation Ilyas, Naveed; Shahzad, Ahsan; Kim, Kiseon Sensors, Vol. 20, Issue 1 https://doi.org/10.3390/s20010043	journal	December 2019

Similar Records

Modular Spiking Neural Circuits for Mapping Long Short-Term Memory on a Neurosynaptic Processor

Journal Article · Thu Jul 12 20:00:00 EDT 2018 · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · OSTI ID:1634300

Statistical Hypothesis Testing using CNN Features for Synthesis of Adversarial Counterexamples to Human and Object Detection Vision Systems

Technical Report · Mon May 01 00:00:00 EDT 2017 · OSTI ID:1361358

Enhancing Nanoparticle Detection in Interferometric Scattering (iSCAT) Microscopy Using a Mask R-CNN

Journal Article · Tue Apr 18 20:00:00 EDT 2023 · Journal of Physical Chemistry. B · OSTI ID:1983732

Related Subjects

97 MATHEMATICS AND COMPUTING
IBM TrueNorth Neurosynaptic System
aerial image analysis
convolutional neural network
deep learning
neuromorphic computing
spiking neural network

REMODEL: Rethinking Deep CNN Models to Detect and Count on a NeuroSynaptic System

Citation Formats

References (25)

Cited By (1)

Similar Records

Related Subjects