New Features for Deep Learning with MATLAB R2017b

The R2017b release of MathWorks products includes many new capabilities for deep learning. Extracted from Mathworks blog by Steve Eddins

New network types and pretrained networks

The heart of deep learning for MATLAB is, of course, the Neural Network Toolbox. The Neural Network Toolbox introduced two new types of networks that you can build and train and apply: directed acyclic graph (DAG) networks, and long short-term memory (LSTM) networks.

In a DAG network, a layer can have inputs from multiple layers instead of just one one. A layer can also output to multiple layers. Here's a sample from the example Create and Train DAG Network for Deep Learning.

You can try out the pretrained GoogLeNet model, which is a DAG network that you can load using googlenet.

Experiment also with long short-term memory (LSTM) networks, which have the ability to learn long-term dependencies in time-series data.

New layer types

There's a pile of new layer types, too: batch normalization, transposed convolution, max unpooling, leaky ReLU, clipped rectified ReLU, addition, and depth concatenation.

My colleague Joe used the Neural Network Toolbox to define his own type of network layer based on a paper he read a couple of months ago. I'll show you his work in detail a little later this fall.

Improvements in network training

When you train your networks, you can now plot the training progress. You can also validate network performance and automatically halt training based on the validation metrics. Plus, you can find optimal network parameters and training options using Bayesian optimization.

Automatic image preprocessing and augmentation is now available for network training. Image augmentation is the idea of increasing the training set by randomly applying transformations, such as resizing, rotation, reflection, and translation, to the available images.

Semantic segmentation

As an image processing algorithms person, I am especially intrigued by the new semantic segmentation capability, which lets you classify pixel regions and visualize the results.

See "Semantic Segmentation Using Deep Learning" for a detailed example using the CamVid dataset from the University of Cambridge.

Deployment to embedded systems

If you are implementing deep learning methods in embedded system, take a look at GPU Coder, a brand new product in the R2017b release. GPU Coder generates CUDA from MATLAB code for deep learning, embedded vision, and autonomous systems. The generated code is well optimized, as you can see from this performance benchmark plot.

MathWorks benchmarks of inference performance of AlexNet using GPU acceleration, Titan XP GPU, Intel® Xeon® CPU E5-1650 v4 at 3.60GHz, cuDNN v5, and Windows 10. Software versions: MATLAB (R2017b), TensorFlow (1.2.0), MXNet (0.10), and Caffe2 (0.8.1).

Here are some additional sources of information.

Leave your comments

Post comment as a guest

0
terms and conditions.

Comments

Course Registration Form


Course Title
Invalid Input

or Key in Your Own Title
Invalid Input

Course Start Date

Invalid Input

Sponsorship (*)
Invalid Input


Contact Person


Salutation(*)
Invalid Input

Name(*)
Invalid Input

Designation/ Department/ Division(*)
Invalid Input

Company(*)
Invalid Input

Billing Address (*)
Invalid Input

Street Address

(*)
Invalid Input

Street Address Line 2

City(*)
Invalid Input

State / Province(*)
Invalid Input

Postal / Zip Code(*)
Invalid Input

Telephone(*)
Invalid Input

Fax
Invalid Input

Email Address (*)
Invalid Input


Participant Details


Participant Salution 1
Invalid Input

Participant Name1
Invalid Input

Designation/ Department/ Division
Invalid Input

Telephone
Invalid Input

Fax
Invalid Input

Email Address
Invalid Input

Dietary Requirement
Invalid Input


Participant Salution 2
Invalid Input

Participant Name2
Invalid Input

Designation/ Department/ Division
Invalid Input

Telephone
Invalid Input

Fax
Invalid Input

Email Address
Invalid Input

Dietary Requirement
Invalid Input


Participant Salution 3
Invalid Input

Participant Name 3
Invalid Input

Designation/ Department/ Division
Invalid Input

Telephone
Invalid Input

Fax
Invalid Input

Email Address
Invalid Input

Dietary Requirement
Invalid Input


Payment Method(*)
Invalid Input

Cheque number
Invalid Input

PO Number
Invalid Input

How did you get to know about this programme?(*)
Invalid Input

Terms and Conditions
Invalid Input

Invalid Input