China’s computer vision unicorns face challenges on many fronts as they expand beyond surveillance: SCMP report
- New report from SCMP Research delves into computer vision unicorns that have made global headlines against the backdrop of US-China tech rivalry
- Companies such as Megvii, SenseTime and Yitu are expanding from surveillance into areas such as retail, health care, finance and more
A new report by the South China Morning Post finds that while the country’s computer vision (CV) industry is growing rapidly, market leaders such as SenseTime, Megvii and Yitu are facing a growing number of challenges, both at home and overseas as tech rivalry increases with the US.
The China AI Deep-Dive: Computer Vision Report, released this week, finds that the main challenge for the country’s computer vision champions is to expand their expertise in this branch of artificial intelligence from mainstays such as surveillance and security into areas such as retail, finance, health care and autonomous vehicles.
Computer vision is a field of computer science that enables computers to visually identify and process objects in digital images or videos. It currently takes up a 50 per cent share of China’s entire AI market, and the country’s CV unicorns lead the world in terms of funding, with US$4.5 billion in venture capital raised between 2016-18.
In the 1970s, US inventor and futurist Raymond Kurzweil presented one of the first commercial products that enabled machines to “see” the real world: the omni-font Optical Character Recognition system, a software program that turned images of text into code. Today, artificial intelligence-powered computer vision is employed in applications ranging from facial recognition to guiding autonomous vehicles.
China has identified computer vision as one of the core areas of AI in its policies, while companies working in the field have benefited from the country’s application of it to surveillance and smart city projects. However, some of China’s CV companies have been added to the US entity list, which blocks access to US-origin components, after Washington alleged that some of their technology had been used to monitor Muslim minority groups in China’s western Xinjiang region.
The Post released its inaugural China AI Report 2020 in February, which examined how government direction, access to large troves of big data, robust funding and entrepreneurial zeal had allowed the country to take a lead in the deployment of artificial intelligence.
The latest report is part of a series taking a deeper look into China’s AI industry. Based on the SCMP technology team’s reporting on the AI sector and data collection by SCMP Research, the China AI Deep-Dive: Computer Vision Report captures how this sector is at the cutting edge of China’s AI drive and rivalry with the US.
“The Computer Vision Report expands the breadth and depth of our tech coverage, diving into China’s largest AI sector to unearth the emerging applications of this ground-breaking technology,” said Tammy Tam, SCMP editor-in-chief. “China is determined to uncover the next big application for computer vision and this report aims to meet that ambition with cutting-edge reporting and analysis to keep the global technology and investment community informed on the latest developments.”
The report provides in-depth case studies of Chinese AI companies that have been pioneering different applications of computer vision, as well as overviews of ecosystems, commercialisation, policies and patents.
It also covers some common questions about the sector – including whether the industry can reduce its dependence on US technology and halt a brain drain.
How China resolves some of these questions and confronts other key challenges, such as privacy concerns and dependence on manually-labeled data, will determine how China’s computer vision sector continues to evolve, finds the report.