Where: Hall 1, Building A
Start Time: 09:00
End Time: 17:30
As part of the Embedded Vision Summit, we’re delighted to have this advanced workshop by Intel® for deep learning experts, computer vision developers, and data scientists with an advanced understanding of machine learning and deep learning techniques who would like to learn optimization and acceleration techniques for industrial, medical, smart city and smart retail applications.
Date: Thursday, May 23, 2019
Time: 9:00 am – 5:00 pm. Badge pickup with coffee and pastries starts at 8:00 am.
Join us for a one-day, hands-on workshop where Intel will take you through advanced topics in accelerating computer vision and deep learning applications using the latest Intel technologies and the OpenVINO™ toolkit. We will walk you through deploying your deep learning application from various frameworks such as Caffe, TensorFlow, and ONNX on different Intel hardware accelerators, CPU, integrated GPU, Movidius NCS, FPGA, etc. using the OpenVINO toolkit. You will also learn how add custom layers for your model to infer it using the OpenVINO toolkit.
- OpenVINO toolkit overview — We will present an overview of the OpenVINO toolkit to do inference on various hardware platforms and discuss its components in detail. We will showcase a demo of framework agnostic features of the OpenVINO toolkit by inferring models from different frameworks such as Caffe, TensorFlow, ONNX.
- Object detection: hands-on lab — We will do a hands-on lab with inferring the YOLO3 model for detecting objects on different hardware such as CPU, integrated GPU, Movidius Compute Stick NCS2 and FPGA.
- Hardware accelerators — Intel FPGA: hands-on lab — We will talk about how FPGA acceleration works, primitives supported in the bitstreams and choosing the right bitstream for your deep learning model. We will do a hands-on lab using the object detection example on Intel FPGA.
- Custom layers: hands-on lab — We will present how to add custom layers for your deep learning model in the OpenVINO toolkit for CPU and the integrated GPU. You will do a hands-on exercise for adding the custom layer with a simple function for the TensorFlow model.
Who will benefit: Deep learning experts, computer vision developers, and data scientists with an advanced understanding of machine learning and deep learning techniques who would like to learn optimization and acceleration techniques for industrial, medical, smart city and smart retail applications.
What you will gain:
- Working knowledge of using the OpenVINO toolkit for inference
- Hands-on application development with the OpenVINO toolkit, including:
- Developing an object detection inference application
- Retargeting the application to use hardware acceleration technologies including GPU, VPU and FPGA
- Adding custom layer for your deep learning model in OpenVINO toolkit for inference
Workshop Sponsor: Intel
Workshop registration is subject to approval by the sponsor.