Success Story: Low-latency AI-based gesture recognition HMI running on edge platforms

NCC-BULGARIA

NCC-Bulgaria is founded by the Institute of Information and Communication Technologies at the Bulgarian Academy of Sciences, the Sofia University “St. Kliment Ohridski” and the University of National and World Economy.

NCC-Bulgaria is focused on:

• Creating a roadmap for successful work in the field of high performance computing, big data analysis and artificial intelligence.
• Analyzing the existing competencies and facilitating the use of HPC/HPDA/AI in Bulgaria
• Raising awareness and promoting HPC/HPDA/AI use in companies and the public sector.

Industrial Organisations Involved:

Software Company Ltd.

Technical/scientific Challenge:

Implement a low-cost, low-latency, high-accuracy gesture recognition solution that can be deployed on different edge platforms such as the NVIDIA Jetson Nano.

Solution:

Using novel approaches to the problem: EAVISE – the final implementation is based. As a result,  a small footprint neural network was created, which consists of CNN  and a multi-stage TCN that allows to work with low-resolution images taken by cheap thermal cameras. This solution was implemented inside the BonsAPPs AIAsset framework, as part of the BonsAPPs challenge. This framework abstracts away the hardware platform on which the AI app runs on allowing us to focus on the AI app instead of hardware specifics.

Business impact:

Since the aim has been to run the optimised neural network on low-power devices, while ensuring both high accuracy and low latency, the final model had to be

well-optimised to run on the selected devices.

That is why we automated the training, quantisation, export and deployment to the testing platform parts of the development process by using gitlab CI/CD pipelines running on a HPC cluster with multiple Nvidia V100 compute units. That way we could iterate much faster since the time between each change and deployment had been significantly shorted by said CI/CD pipelines.

Using a HPC cluster also allowed the team to work with much larger datasets that a regular workstation computer would not even be able to load into memory.

Benefits:

The  Software Company Ltd. was recognised by the BONSAPPS as a reliable partner due to support by our team, resulted in a solution of an important problem using cutting-edge scientific research: implementing  HPC technology to improve both the development process and the quality of the final AI application.