Hands-on Session Programme (Tentative)

Overview

The hands-on session will be supported by the OpenCV China Team. Participants will learn to develop a real-time face recognition system on the EAIDK-310 development board. All participants will receive an ARM development board, EAIDK-310, free of charge. The three teams with best works will be awarded a more advanced EAIDK-610 development kits. Thanks to OPEN AI LAB for sponsoring EAIDK-310 and EAIDK-610 development boards.

To finish the project, the participants should form a team. There should be 2 participants from different affiliations in each team. The team can also choose topics other than face recognition but should get approval from instructors. The development boards will be mailed to online participants before the winter school.

What should participants prepare?
1. A laptop with a network cable interface.
2. Install VNC Viewer on laptop.

EAIDK-310 development board

25 Jan (Day 2)

13:30 – 14:00

Topic: the new features of OpenCV 4.x and 5

Speaker: Vadim Pisarevsky, OpenCV China Team
Slides

14:00 – 14:30

Topic: How to develop a face recognition system on an ARM dev board using OpenCV

Speaker: Jia Wu, OpenCV China Team
Slides

14:30 – 15:30

Students develop a biometric system with the help of instructors. And students can keep developing the system after the session with the support of instructors and submit the final technical report on 27 Jan, 2021. The organizer will score these reports. And the top 3 teams will be awarded advanced EAIDK-610 development kits.

26 Jan (Day 3)

13:30 - 15:30

Students continue to develop a biometric system with the help of instructors. Students can keep developing the system after the session with the support of instructors.

27 Jan (Day 4)

13:30 - 15:30

Students continue to develop a biometric system with the help of instructors. Note that, don’t forget to submit the final technical report and demo of your project to jia.wu@opencv.org.cn before 23:59 Jan 27, 2021.

28 Jan (Day 5)

14:30 - 15:30

Final prize presentation + presentation (mixed mode)
The three awarded teams should present their work. (30 minutes per team)

EAIDK-310 Specifications

EAIDK-310 produced by OPEN AI LAB provides an embedded AI development platform, with Linux and embedded deep learning framework Tengine which supports the deployment of Caffe/TensorFlow/Pytorch/MxNet/ONNX/Darknet pre-installed. The default system is based on Fedora 28. It is similar with Raspberry Pi but easier to use.