Bringing neural networks to cellphones

MIT researchers have designed new methods for paring down neural networks so that they’ll run more efficiently on handheld devices.

08/04/2017


In recent years, the best-performing artificial-intelligence systems, in areas such as autonomous driving, speech recognition, computer vision, and automatic translation, have come courtesy of software systems known as neural networks.

But neural networks take up a lot of memory and consume a lot of power, so they usually run on servers in the cloud, which receive data from desktop or mobile devices and then send back their analyses.

MIT researchers have designed new methods for paring down neural networks so that they’ll run more efficiently on handheld devices. Courtesy: MIT; Jose-Luis Olivares, photographerIn 2016, MIT associate professor of electrical engineering and computer science Vivienne Sze and colleagues unveiled a new, energy-efficient computer chip optimized for neural networks, which could enable powerful artificial-intelligence systems to run locally on mobile devices.

Now, Sze and her colleagues have approached the same problem from the opposite direction, with a battery of techniques for designing more energy-efficient neural networks. First, they developed an analytic method that can determine how much power a neural network will consume when run on a particular type of hardware. Then they used the method to evaluate new techniques for paring down neural networks so that they'll run more efficiently on handheld devices.

The researchers describe the work in a paper they presented at the Computer Vision and Pattern Recognition Conference. In the paper, they report that the methods offered as much as a 73% reduction in power consumption over the standard implementation of neural networks, and as much as a 43% reduction over the best previous method for paring the networks down.

Energy evaluator

Loosely based on the anatomy of the brain, neural networks consist of thousands or even millions of simple but densely interconnected information-processing nodes, usually organized into layers. Different types of networks vary according to their number of layers, the number of connections between the nodes, and the number of nodes in each layer.

The connections between nodes have "weights" associated with them, which determine how much a given node's output will contribute to the next node's computation. During training, in which the network is presented with examples of the computation it's learning to perform, those weights are continually readjusted, until the output of the network's last layer consistently corresponds with the result of the computation.

"The first thing we did was develop an energy-modeling tool that accounts for data movement, transactions, and data flow," Sze said. "If you give it a network architecture and the value of its weights, it will tell you how much energy this neural network will take. One of the questions that people had is, 'Is it more energy efficient to have a shallow network and more weights or a deeper network with fewer weights?' This tool gives us better intuition as to where the energy is going, so that an algorithm designer could have a better understanding and use this as feedback. The second thing we did is that, now that we know where the energy is actually going, we started to use this model to drive our design of energy-efficient neural networks."

In the past, Sze explained, researchers attempting to reduce neural networks' power consumption used a technique called "pruning." Low-weight connections between nodes contribute very little to a neural network's final output, so many of them can be safely eliminated, or pruned.

Principled pruning

With the aid of their energy model, Sze and her colleagues, Tien-Ju Yang and Yu-Hsin Chen, both graduate students in electrical engineering and computer science, varied this approach. Although cutting even a large number of low-weight connections can have little effect on a neural net's output, cutting all of them probably would, so pruning techniques must have some mechanism for deciding when to stop.

The MIT researchers thus begin pruning those layers of the network that consume the most energy. That way, the cuts translate to the greatest possible energy savings. They call this method "energy-aware pruning."

Weights in a neural network can be either positive or negative, so the researchers' method also looks for cases in which connections with weights of opposite sign tend to cancel each other out. The inputs to a given node are the outputs of nodes in the layer below, multiplied by the weights of their connections. So the researchers' method looks not only at the weights but also at the way the associated nodes handle training data. Only if groups of connections with positive and negative weights consistently offset each other can they be safely cut. This leads to more efficient networks with fewer connections than earlier pruning methods did.

Massachusetts Institute of Technology

www.mit.edu 

- Edited by Hannah Cox, content specialist, Control Engineering, CFE Media, hcox@cfemedia.com. See more Control Engineering manufacturing IT stories.



Engineers' Choice Awards
The Engineers' Choice Awards highlight some of the best new control, instrumentation and automation products as chosen by Control Engineering subscribers.
System Integrator Giants
The System Integrator Giants program lists the top 100 system integrators among companies listed in CFE Media's Global System Integrator Database.
System Integrator of the Year
Each year, a panel of Control Engineering and Plant Engineering editors and industry expert judges select the System Integrator of the Year Award winners in three categories.
Design of Safe and Reliable Hydraulic Systems for Subsea Applications
This eGuide explains how the operation of hydraulic systems for subsea applications requires the user to consider additional aspects because of the unique conditions that apply to the setting
How to Maximize Factory Automation Efficiency with Low Cost Machine Vision
This eGuide illustrates solutions, applications and benefits of machine vision systems.
Wireless Reliability in Harsh Environments
Learn how to increase device reliability in harsh environments and decrease unplanned system downtime.
October 2018
HMI hardware evolution, Data acquisition strategies, Matching motors and drives, Machine vision advice
September 2018
Optimize controls via cloud software, ladder logic simulation, industrial wireless best practices
August 2018
Augmented reality and virtual reality education, autotuning PID control, cybersecurity advice, educating engineers
Edge Computing
This article collection contains several articles on how today's technologies heap benefits onto an edge-computing architecture such as faster computing, better networking, more memory, smarter analytics, cloud-based intelligence, and lower costs.
Data Center Design
Data centers, data closets, edge and cloud computing, co-location facilities, and similar topics are among the fastest-changing in the industry.
IIoT: Machines, Equipment, & Asset Management
Articles in this digital report highlight technologies that enable Industrial Internet of Things, IIoT-related products and strategies.
SIDB

Find and connect with the most suitable service provider for your unique application. Start searching the Global System Integrator Database Now!

October 2018
2018 Product of the Year; Subsurface data methodologies; Digital twins; Well lifecycle data
August 2018
SCADA standardization, capital expenditures, data-driven drilling and execution
June 2018
Machine learning, produced water benefits, progressive cavity pumps
John O. Ayuk, PE, CFSE, PMP, CAP
Automation Engineer; Wood Group
Doug Baker
System Integrator; Cross Integrated Systems Group
Jose S. Vasquez, Jr.
Jose S. Vasquez, Jr.
Fire & Life Safety Engineer; Technip USA Inc.
Data Centers: Impacts of Climate and Cooling Technology
This course focuses on climate analysis, appropriateness of cooling system selection, and combining cooling systems.
Safety First: Arc Flash 101
This course will help identify and reveal electrical hazards and identify the solutions to implementing and maintaining a safe work environment.
Critical Power: Hospital Electrical Systems
This course explains how maintaining power and communication systems through emergency power-generation systems is critical.
Engineers' Choice Awards
The Engineers' Choice Awards highlight some of the best new control, instrumentation and automation products as chosen by Control Engineering subscribers.
System Integrator Giants
The System Integrator Giants program lists the top 100 system integrators among companies listed in CFE Media's Global System Integrator Database.
System Integrator of the Year
Each year, a panel of Control Engineering and Plant Engineering editors and industry expert judges select the System Integrator of the Year Award winners in three categories.
Design of Safe and Reliable Hydraulic Systems for Subsea Applications
This eGuide explains how the operation of hydraulic systems for subsea applications requires the user to consider additional aspects because of the unique conditions that apply to the setting
How to Maximize Factory Automation Efficiency with Low Cost Machine Vision
This eGuide illustrates solutions, applications and benefits of machine vision systems.
Wireless Reliability in Harsh Environments
Learn how to increase device reliability in harsh environments and decrease unplanned system downtime.
October 2018
HMI hardware evolution, Data acquisition strategies, Matching motors and drives, Machine vision advice
September 2018
Optimize controls via cloud software, ladder logic simulation, industrial wireless best practices
August 2018
Augmented reality and virtual reality education, autotuning PID control, cybersecurity advice, educating engineers
Edge Computing
This article collection contains several articles on how today's technologies heap benefits onto an edge-computing architecture such as faster computing, better networking, more memory, smarter analytics, cloud-based intelligence, and lower costs.
Data Center Design
Data centers, data closets, edge and cloud computing, co-location facilities, and similar topics are among the fastest-changing in the industry.
IIoT: Machines, Equipment, & Asset Management
Articles in this digital report highlight technologies that enable Industrial Internet of Things, IIoT-related products and strategies.
SIDB

Find and connect with the most suitable service provider for your unique application. Start searching the Global System Integrator Database Now!

October 2018
2018 Product of the Year; Subsurface data methodologies; Digital twins; Well lifecycle data
August 2018
SCADA standardization, capital expenditures, data-driven drilling and execution
June 2018
Machine learning, produced water benefits, progressive cavity pumps
John O. Ayuk, PE, CFSE, PMP, CAP
Automation Engineer; Wood Group
Doug Baker
System Integrator; Cross Integrated Systems Group
Jose S. Vasquez, Jr.
Jose S. Vasquez, Jr.
Fire & Life Safety Engineer; Technip USA Inc.
Data Centers: Impacts of Climate and Cooling Technology
This course focuses on climate analysis, appropriateness of cooling system selection, and combining cooling systems.
Safety First: Arc Flash 101
This course will help identify and reveal electrical hazards and identify the solutions to implementing and maintaining a safe work environment.
Critical Power: Hospital Electrical Systems
This course explains how maintaining power and communication systems through emergency power-generation systems is critical.
Engineers' Choice Awards
The Engineers' Choice Awards highlight some of the best new control, instrumentation and automation products as chosen by Control Engineering subscribers.
System Integrator Giants
The System Integrator Giants program lists the top 100 system integrators among companies listed in CFE Media's Global System Integrator Database.
System Integrator of the Year
Each year, a panel of Control Engineering and Plant Engineering editors and industry expert judges select the System Integrator of the Year Award winners in three categories.
Design of Safe and Reliable Hydraulic Systems for Subsea Applications
This eGuide explains how the operation of hydraulic systems for subsea applications requires the user to consider additional aspects because of the unique conditions that apply to the setting
How to Maximize Factory Automation Efficiency with Low Cost Machine Vision
This eGuide illustrates solutions, applications and benefits of machine vision systems.
Wireless Reliability in Harsh Environments
Learn how to increase device reliability in harsh environments and decrease unplanned system downtime.
October 2018
HMI hardware evolution, Data acquisition strategies, Matching motors and drives, Machine vision advice
September 2018
Optimize controls via cloud software, ladder logic simulation, industrial wireless best practices
August 2018
Augmented reality and virtual reality education, autotuning PID control, cybersecurity advice, educating engineers
Edge Computing
This article collection contains several articles on how today's technologies heap benefits onto an edge-computing architecture such as faster computing, better networking, more memory, smarter analytics, cloud-based intelligence, and lower costs.
Data Center Design
Data centers, data closets, edge and cloud computing, co-location facilities, and similar topics are among the fastest-changing in the industry.
IIoT: Machines, Equipment, & Asset Management
Articles in this digital report highlight technologies that enable Industrial Internet of Things, IIoT-related products and strategies.
SIDB

Find and connect with the most suitable service provider for your unique application. Start searching the Global System Integrator Database Now!

October 2018
2018 Product of the Year; Subsurface data methodologies; Digital twins; Well lifecycle data
August 2018
SCADA standardization, capital expenditures, data-driven drilling and execution
June 2018
Machine learning, produced water benefits, progressive cavity pumps
John O. Ayuk, PE, CFSE, PMP, CAP
Automation Engineer; Wood Group
Doug Baker
System Integrator; Cross Integrated Systems Group
Jose S. Vasquez, Jr.
Jose S. Vasquez, Jr.
Fire & Life Safety Engineer; Technip USA Inc.
Data Centers: Impacts of Climate and Cooling Technology
This course focuses on climate analysis, appropriateness of cooling system selection, and combining cooling systems.
Safety First: Arc Flash 101
This course will help identify and reveal electrical hazards and identify the solutions to implementing and maintaining a safe work environment.
Critical Power: Hospital Electrical Systems
This course explains how maintaining power and communication systems through emergency power-generation systems is critical.
click me