Hands-on MATLAB Workshop
Designing and deploying deep learning based computer vision applications to embedded CPU and GPU platforms is challenging because of resource constraints inherent in embedded devices. A MATLAB® based workflow facilitates the design of these applications, and automatically generated C or CUDA® code can be deployed on embedded boards to achieve very fast inference. The workshop illustrates how MATLAB supports all major phases of this workflow. Starting with algorithm design, the algorithm may employ deep neural networks augmented with traditional computer vision techniques and can be tested and verified within MATLAB. Next, these networks are trained using GPU and parallel computing support for MATLAB either on the desktop, cluster, or the cloud.
Introduction to Deep Learning for Computer Vision Applications Using MATLAB
- Use a pre-trained network for image classification
- Build a deep learning network from scratch
- Perform transfer learning
Addressing Challenges in Deep Learning Workflows Using MATLAB
- Accelerating the labelling process using automation algorithms
- Understanding network behaviour using visualizations
The presenters will also discuss the following topics during the workshop:
● Labelling large amount of images
● Hyperparameter tuning of deep neural networks
● Scaling up training to GPUs, multi-GPUs and clusters
● Deployment workflows for desktop, web, and cloud
● Deployment workflows using automatic code generation for embedded platforms
(NVIDIA GPU, Intel CPU, ARM CPU)
PRE-REQUISITES FOR HANDS-ON WORKSHOP:
Laptops with MATLAB (version R2017b or later) installed. Preferably R2018b (latest).
The list of the toolboxes needed for the workshop includes:
MATLAB® and a full set of products for deep learning: Neural Network Toolbox™, Statistics and
Machine Learning Toolbox™, Parallel Computing Toolbox™, Image Processing Toolbox™,
Computer Vision System Toolbox™, Image Acquisition Toolbox™, and Signal Processing
Ones the MATLAB is up and running Download and Install pre-trained network Alexnet(download)
All names and Trademarks belong to respective companies.