Workshop: Maximize AI performance through computer vision based innovative retail solution - Automated Self-Checkout

With the growth of AI and cloud technologies it's no surprise that many have tried to move their compute to the cloud. On paper this sounds like a perfect solution but in the real world it presents many challenges one of those being cost. In particular computer vision frame data can be very large and expensive to compute. By taking advantage of local compute devices you can filter out unnecessary frames and significantly reduce costs. But where do you start and how do you know how much compute is needed? The https://github.com/intel-retail/automated-self-checkout open source project provides the tools needed to launch and benchmark a computer vision based workload on your local device. In this workshop we will demonstrate how to utilize the automated-self-checkout solution to run Docker containers with open source Intel OpenVINO toolkit and benchmark their performance on different hardware platforms. Through this hands-on workshop, attendees will be able to setup and run the AI computer vision based retail pipelines using the Automated Self-Checkout reference implementation solution. Attendees will also gain knowledge on how to benchmark the workloads using this tool for obtaining maximum hardware performance. Though this project can work on any intel based system (6th gen core or above) with ubuntu (20.04 or 22.04), we will provide the hardware setup required to run this workshop smoothly.

Please Note: Presenters for this workshop will be from my team, they were unable to submit the abstract due to bug with the submission tool