asmag logo
Hikvision: What to expect from video in the age of AI?
the a&s Editorial Team 2017/12/6

a&s talks to Keen Yao, VP of International Business Center at Hikvision Digital Technology about how artificial intelligence (AI) technologies are being implemented in many different verticals.According to a new report from Tractica, “AI is an umbrella term that includes multiple technologies, such as machine learning, deep learning, computer vision, natural language processing (NLP), machine reasoning and strong AI.” Machine learning and deep learning are two important and effective AI technologies for security-related applications, and are able to provide many extra values, such as creating business value, saving labor costs and increasing productivity for an enterprise user. The report also predicted that AI software revenue will grow from US$1.4 billion in 2016 to $59.8 billion by 2025.

Hikvision also sees the high business potential of using AI in a video surveillance system, Keen Yao, VP of International Business Center at Hikvision Digital Technology, commented, “AI technology deployment is now in its fastest-growing stage in many sectors. It expands the variety of applications for video surveillance. When AI is coupled with big data analysis, a lot of business values can be created. Many of our users have been expressing a high interest for AI.”

China can be considered as one of the countries that embraces facial recognition technology the most — people use it to gain access to doors and make transactions, while retailers can recommend (menu) items based on factors including age, gender and mood by scanning customers’ faces. The Chinese government is also a big fan, applying facial recognition extensively in their surveillance cameras for public safety

AI future development

Growing in the country with a great emphasis on the AI technology, Yao said, “AI is a very important element for Hikvision’s systems. The company has invested in video intelligence technology since 2006, when its first algorithm team was established. In 2013, the company started to implement deep learning technology, and in 2016, our full range of deep learning products were introduced to the security industry.”

To further optimize AI performance in video surveillance systems, Yao said, “There should be even closer future collaboration with key component chipset companies. Due to this huge potential, we may see AI-enabled SoC (system on a chip) for video surveillance systems on the market in the future.”

In addition to key component development, computing and storage are also two major issues related to AI performance. Hikvision’s “AI Cloud” establishes a distributed structure incorporating cloud computing and edge computing. It extends the AI algorithm from the cloud to an edge network of on-premises video recorders and servers, and further to edge devices like security cameras. The three layer architecture supports the development of a new class of AI-powered application, becoming even smarter and faster.

AI use cases

Yao highlighted several AI use cases. He mentioned how Hikvision has helped the Haining Public Safety Bureau in China to build a public safety information system based on a big data analytics infrastructure. The bureau now relies on this new technology to transform police roles and services. Without increasing their staff, the bureau was still able to become more productive as the information system helps them deal with traffic monitoring, solve many criminal cases and enhance overall law enforcement effectiveness.

Another application case showed how Hikvision’s deep learning panoramic cameras with advanced behavior analysis are being leveraged by a football club to clearly analyze game tactics used by sports players and their adversaries.

Be innovative

AI has turned many possibilities into near realities. Hence, we see an emerging interest for AI from end users and this is driving current technology developments. Security companies would need to stay innovative to stay competitive in the market.

Messe Frankfurt New Era Business Media Ltd. All rights reserved. 2019/10/17 print out