** To register for this workshop please start the CEvent registration process for IEEE GEM conference; when you get to the 4th screen you can register for workshops, without the main conference. For additional information contact firstname.lastname@example.org
Date & Venue: 09.00 – 15.30, 15th August 2018; Alice Perry Engineering Building, NUI Galway
Deep learning frameworks have reached the point where they can be used as tools in embedded electronics devices, allowing the production of “smart” products that are capable of recognizing objects and making decisions. This workshop introduces participants to the basics of practical deep learning, as well as an understanding of approaches and options for solving real world problems. In addition participants will have the opportunity to practice developing and deploying artificial intelligence models with the latest NVIDIA GPU’s, Raspberry PI and Movidus USB sticks using Python, Tensorflow, and Caffe. The workshop will include 2 hands-on lab sessions.
Researchers, engineers, computer programmers, technical managers & project leads who want to understand what deep learning is and how it can be used to bring intelligence to software and products, especially embedded systems and Internet-of-Things (“Edge”) devices.
- A basic knowledge of Python or equivalent scripting language.
- Some general programming experience.
- A basic knowledge of unix/linux or equivalent command line interfaces.
- Participants will gain an understanding of the “pipeline” of a deep learning project, from training to deployment of a model on a device as well as the hardware and software required.
- Participants will be introduced to the abilities and limits of the latest techniques in artificial intelligence as applied to a selection of computer vision tasks.
|Peter Corcoran is a Fellow of IEEE, past Editor-in-Chief of IEEE Consumer Electronics Magazine and a Professor at NUI Galway. His research interests include biometrics, imaging, deep learning, edge-AI and consumer electronics. He is co-author on 350+ technical publications and co-inventor on more than 300 granted US patents. In addition to his academic career, he is an occasional entrepreneur, industry consultant and compulsive inventor.
|Shabab Bazrafkan received a B.Sc. degree in electrical engineering from Urmia University, Urmia, Iran, in 2011, and the M.Sc. degree in telecommunication engineering, image processing branch from the Shiraz University of Technology in 2013. He is currently pursuing the Ph.D. degree with the National University of Ireland Galway. He is currently with FotoNation, Ltd. His field of working is deep neural networks and neural network design.
|Joseph Lemley received a B.S. degree in computer science and the master’s degree in computational science from Central Washington University in 2006 and 2016, respectively. He is currently pursuing the Ph.D. with the National University of Ireland Galway. His field of work is machine learning using deep neural networks for tasks related to computer vision. His Ph.D. is funded by FotoNation, Ltd., under the IRCSET Employment Ph.D. Program.
|Introduction to Deep Learning & Edge-AI|
|09.00||Deep Learning 101 – What is it and why is it so important?||Peter Corcoran & Joseph Lemley|
|09.30||Hardware & Software basics for DL||Joseph Lemley & Shabab Bazrafkan|
|10.15||Deep Learning Frameworks & ready-made networks||Shabab Bazrafkan & Joseph Lemley|
|Hands-on #1 – Practical DL with a GPU cluster|
|11.30||A Crash course in Tensorflow||Joseph Lemley & Shabab Bazrafkan|
|12.15||Hands-on Challenge with Tensorflow on GPU cluster||Joseph Lemley, Shabab Bazrafkan, Viktor Varakaris & Joe Desbonnet|
|13.15||Sandwich Lunch & continue with challenge|
|Hands-on #2 – Edge-AI with Raspberry PI|
|13.45||Crash course in Raspberry PI & Movidius AI-stick||Joseph Lemley & Adrian Ungureanu|
|14.30||Hands-on Challenge with Yolo & Squeezenet running on RPI and Movidius AI-stick.||Joseph Lemley, Aoife McDonagh, Adrian Ungureanu & Joe Desbonnet|
|16.00||Wrap-Up & Coffee Break|
Lunch & Coffee provided
|Booked with IEEE GEM Conference||Stand-alone||At Conference|