HPRC/Intel OpenVINO Workshop
Time: Friday March 6, 2020 - 9:00am-4:00pm
Location: SCC 102.B
Instructor: Luke Ge, Intel
Register for Workshop here
This workshop provides easy access to the fundamental concepts of the Intel Distribution of OpenVINO toolkit for Computer Vision Applications. Throughout this course, you will be introduced to demos, showcasing the capabilities of this toolkit.
Session Name | Description | Time | Duration |
---|---|---|---|
Introduction of AI and Intel AI portfolio | This session will give an introduction of AI/Deep Learning (e.g. training and inferencing) and a quick, high-level overview of Intel AI technologies. | 9:00 - 9:20am |
20 min |
AI Inference: using Intel OpenVINO SDK | This session will focus on OpenVINO™ - Intel's open source toolkit for Inference. OpenVINO™ is used to accelerate and deploy pre-trained workloads across the entire suite of Intel® hardware (including accelerators). | 9:20 - 10:20am |
60 min |
Break/Network | 10:20 - 10:30am |
10 min | |
Hands-on sessions –Inference | Attendees will have a chance to try out various Intel® AI solutions hands-on. Attendees will be walked through setup of remote access to a Jupyter Notebook based learning of optimizing deep learning code running on various Intel platform. | 10:30 - 12:00pm |
90 min Luke |
Lunch Break/Network - Sponsored by TAMU FIDS. | 12:00 - 13:00pm |
60 min | |
AI Training and Software Optimization | In this session the speaker will walk through training techniques (like data parallelism and model parallelism) and some software optimization techniques. | 13:00 - 13:30pm |
30 min Luke |
Hands-on sessions – DL Training | Attendees will be walked through a Jupyter Notebook based learning of optimizing deep learning code running on Intel Xeon® Scalable Processor. | 13:30 - 14:30pm |
1 hour |
Break/Network | 14:30 - 14:45pm |
15 min | |
Use cases of large image processing | In this session the speaker will share real-world solutions of processing high resolution, large, 3D images and the techniques to improve model accuracy. | 14:45 - 15:15pm |
30 min Luke |
Live demos | Intel® Solution experts will demonstrate the various topics discussed over the day. We will run live examples on Intel® OpenVINO SDK for inference, distributed training on multi-nodes, Brain tumor detection, Intel ® Movidius for inference. | 15:15 - 15:30pm |
15 min Luke |
Speaker Bio
Luke Ge is an artificial intelligence (AI) Technical Solution Specialist (TSS) of Intel. His particular focus is on AI workload (training, inference) optimization using Intel tools (MKL-DNN, OpenVINO, Intel Tensorflow, etc.) on Intel platform (Xeon, VPU, FPGA, Optane, Omni-Path).