INTERVIEW WITH MR HU YANGZHONG, PRESIDENT, HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO., LTD.

Serving diverse vertical markets covering more than 150 countries, with over 26,000 employees worldwide, and sales revenues up from RMB 31.93 billion in 2016 to RMB 41.91 billion in 2017, there is no disputing that Hikvision ranks as the number 1 supplier of innovative video surveillance products and solutions for 3 consecutive years by a&s Security 50. Team MalaysiaSAFE had the honor of meeting-up with the company’s President, Mr Hu YangZhong when Hikvision hosted the AI Cloud Ecosystem Summit in Malaysia. Some excerpts from his interview are carried below:

 

Mr. Hu Yangzhong, President, Hangzhou Hikvision Digital Technology Co., Ltd.

MalaysiaSAFE: On a personal level, what does it mean to you that Hikvision is an innovative technology company and the world’s leading video security products supplier? Or is that all just business?
Mr. Hu: When I started, my objective was to lead the company and put the business on a growth trajectory. Over the years, I felt an increasingly personal responsibility over the organisation - especially given the company’s passion for constant technological innovation. It is what drives me to keep moving things forward in my pursuit for excellence. And I feel lucky that I have had the opportunity to be part of Hikvision’s growth and success over the years.
MalaysiaSAFE: What is your vision when it comes to what a camera can and should be able to do?
Mr. Hu: Traditionally, security cameras were only used to live-monitor a particular event to deter crime. Then, it became common practice for law-enforcement and businesses to install security cameras since video recording can be used as evidence of criminal activity, and it was a more cost-effective way to deter crime compared to dedicating human resources and security personnel.
But today, cameras do more than just help prevent crime. Cameras with AI technology can be used in hundreds of different IoT applications and industries, and I hope this trend continues where cameras will be integrated into realms that are helpful to our lives.
MalaysiaSAFE: When did the company realise that AI technology or machine learning was going to shape the security industry?
Mr. Hu: In early 2006, Hikvision was using the traditional algorithm embedded in its DVRs, providing some smart functions. Since 2012, Deep Learning technology has been gradually becoming mature, and Hikvision realised the importance of machine learning and AI technologies, so we started investing more on the development of our own Deep Learning algorithm and technology for our systems. In 2015 and 2016, we began to launch our AI enabled hardware and products.

DEEP LEARNING

Deep Learning is a concept that is fundamentally different than machine learning, which operates on traditional intelligent algorithms. It incorporates unsupervised or “self-learning” principles made possible by today’s rapid development of GPUs, cloud computing and high performance hardware. In Deep Learning, as data passes through layers of processing, the machine takes a partial understanding of the features in the object. Then, the computer will extract as many features from the object as possible, until finally recognising it. The more features there are, the more accurate the recognition and classifications will be.
Hikvision takes the Deep Learning concept and applies it to the security industry. The company is developing this concept in its own analytics algorithms. Enhanced accuracy is the result of multi-layer learning based on a large amount of data. Application of this algorithm into face recognition, vehicle recognition, human recognition, and other platforms will significantly advance the performance of analytics. Now, Hikvision’s DeepinView Cameras and DeepinMind NVRs lead the way in the world of surveillance technology by making invisible intelligence visible for users, and then putting that intelligence to good use.
MalaysiaSAFE: How would you rate your business opportunities within this region?
Mr. Hu: It is always a challenge for manufacturers to find a proper way to meet the real needs of the local market. In the Malaysian market, I think security companies should continue to localise the products as well as the services. In the coming years, we will be more focused on localisation, where we can provide better products and services for our customers. We have invested a lot of resources to establish our Malaysian subsidiary, and we look forward to receiving positive feedback.
MalaysiaSAFE: What do you think would be the key obstacles that need to be reconciled before AI technology can reach its full potential?
Mr. Hu: There needs to be a focus on 3 basic sectors: Computing capability, algorithms and data-mining. Plus, more training is needed to develop a talented pool of technical workers who can understand machine learning and enable the optimal use of AI technologies.
MalaysiaSAFE: Are there any plans to jump into mainstream consumer technology industry such as incorporating Hikvision’s Deep Learning algorithms into smartphones?
Mr. Hu: No, our focus will remain on video-based IoT. Besides, Hikvision has a broader portfolio than security cameras. For example, manufacturing plants can use our Machine Vision products to improve their efficiency.
MalaysiaSAFE: What new initiatives has HIKVISION chartered for 2019?
Mr. Hu: We aim to facilitate AI technology into more vertical applications through our AI cloud ecosystem. We also plan to explore other ways where AI technologies can solve real-world problems beyond the security. For example, we are currently engaged with local partners to protect wildlife and some endangered species, including the panda and the bar-headed geese.

AI CLOUD

Hikvision's 'AI Cloud' is a customised architecture of intelligent IoT that incorporates Cloud Computing and Edge Computing. It extends the Artificial Intelligence (AI) algorithm from the cloud, to an edge network of on-premise video recorders and servers, and further to edge devices like the security cameras. It distributes the overall pressure of processing large amounts of data from an organisation’s cloud computing structure. The three layer architecture provides the perfect platform for building a new class of AI-powered solutions such as smart city, public security, smart transportation, intelligent manufacturing and smart retail.

Features