Kingston University has recently won a new Knowledge Transfer Partnership (KTP) worth £150,000 with VCA Technology Ltd. VCA is a UK company that specialises in video analytics software in the areas of retail customer behaviour, security and traffic monitoring. Their highly acclaimed products have been successfully deployed worldwide in sectors ranging from sports events, to the oil and gas industry.
The company has been a leader in embedded video analytics since the release of their first product in 2010. However, due to the 'siliconisation' of CCTV camera functions such as video compression, the amount of processing power available in traditional IP cameras for video analytics has stagnated. Correspondingly, the performance of video analytics algorithms has not kept pace with that of other camera functions.
Driven by the demands of autonomous vehicles, machine learning and artificial intelligence, recent advances in embedded parallel processors such as graphics processing units (GPUs), now offer cost effective deployment of a new generation of video analysis algorithms based on deep learning. The company saw this as an opportunity to develop and embed new, state-of-the-art technology.
VCA aims to develop a more accurate method of counting people in very busy scenes, for example customers in a shop, with standard CCTV cameras. This will involve the use of deep learning and recent developments in hardware to develop a cost-effective, embedded device.
Deep learning is a subfield of machine learning, a type of artificial intelligence that provides computers with the ability to learn without being explicitly programmed. It is concerned with algorithms inspired by the structure and function of the brain called artificial neural networks.
As the subject of deep learning is still in its infancy and most of the existing knowledge resides within academia, it was not possible for VCA to simply send current staff on training courses, or to outsource the work as a consultancy project. This is where a KTP was an ideal solution.
A graduate will be employed to manage the two-year project at VCA's European office, near Kingston. This 'associate' will be managed and supported by both the company and experts in computer science and mathematics at Kingston University.
The aim of the KTP is to enable VCA to develop improved algorithms with better performance. The innovative technology created will provide more accurate results compared to existing approaches and this will allow the company to increase their competitive advantage and market share.
This is directly aligned to one of the strategic aims of the company which is to be a leader in the video analytics market, translating directly to increased licence sales.
"VCA Technology is excited to be partnering with Kingston University to leverage their expertise in deep learning. This partnership will help VCA Technology to develop a new range of more accurate intelligent vision products that take advantage of recent developments in massively parallel central processing units (CPUs) and graphics processing units (GPUs)." Ben White, Chief Technology Officer and Co-Founder, VCA Technology.