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

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

doi:10.3389/fnins.2019.00004

Abstract

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.

Highlights

Neural networks today are achieving state-of-the-art performance in competitions across a range of fields
This dataset is useful for training Deep Neural Networks (DNNs) so that they are able to perform area based surveillance by detecting and counting cars that are present in the image
The efficient deep neuromorphic networks (EEDN)-trained convolutional neural networks (CNNs) structures have been compared against more standard neural network models that were deployed on Titan X GPU

Summary

Introduction

Neural networks today are achieving state-of-the-art performance in competitions across a range of fields. Hardware that mimics the computational capabilities of a human brain through spiking neural networks has been shown to be extremely energy-efficient, and capable of scaling up to large neural networks. One of the major challenges that these spiking neural network-based platforms faced was deploying convolutional neural networks (CNNs) on spiking neurons. This issue was addressed in the recent work from Cao et al (2015) and Esser et al (2016), and Eta Compute (Moore, 2018). The authors in Esser et al (2016) have proposed an algorithm named energy-efficient deep neuromorphic networks (EEDN) to map CNNs on TrueNorth. EEDN networks achieved at or near state of the art accuracy when compared with traditional 32-bit precision neural networks

Objectives

Methods

Results

Conclusion