The deal is Accenture Venture’s first investment in China, and falls under the company’s Open Innovation mantra, which effectively brings its clients closer to young companies working on big ideas, ideas in which Accenture will then have a stake. The value of the investment was not made public.
Already, the focus of the deal – which both sides refer to as an alliance – is on developing joint industry solutions and go-to-market activities. Malong’s core technology is called ProductAI, and is an image-recognition and auto-tagging system with retail applications.
Retailers, they write, can use the technology’s seeing capability to make checkouts more efficient and aid companies working on visual commerce, which is to say shopping by taking a picture with a smartphone.
But the technology goes deeper and forms part of a unified computer vision offer. A cornerstone of Accenture’s computer vision capabilities, “It will provide services and solutions to organizations in any industry where objects need to be reliably recognized at a high-level, microscopic-level, or x-ray level.”
Other use cases include defect detection in manufacturing to drive quality, baggage scanning for improved safety, and analyzing medical images to help doctors make critical decisions faster.
The technology works through a “weakly-supervised” deep learning algorithm, which allows it to train with unusually large data sets. It boasts a 94.78% recognition rate – humans, for comparison, recognised between 94 and 94.9% on the same task.
“Gaining Accenture as an investor and alliance partner will allow us to jointly bring our services to top-tier companies worldwide,” said Malong Technologies’ CEO, Dr. Dinlong Huang.
Computer vision has a long pedigree in the Artificial Intelligence field. According to a Gartner report from May, which profiled Malong, “Computer vision (CV) is virtually synonymous with the development of artificial intelligence since the invention of the perceptron 61 years ago, and CV today stands at the cutting edge of deep-learning implementations.”
While Google, with its Lens product, has been pushing visual search for some time, it is Pinterest that has pioneered an effective consumer application of visual search technology, as a WARC Trend Snapshot details. The technology it gained enabled the social network to launch Pinterest Lens in February 2017. By using the camera within the Pinterest app, users can capture an image in the real world, and then tap to discover similar images and designs on the web.
The technology, aside from having powerful implications for driving revenue and making the step between social media platforms and shopping a little shorter, has also brought researchers expanded capabilities. In Latin America, researchers used computer vision to estimate socioeconomic levels from satellite imagery.
In the last five years, Accenture China has shifted its focus away from the Western tech firms, favouring Chinese businesses that allow the consultancy to help clients into China’s innovation ecosystem. “We are trying to be local in the Chinese market as much as we can,” Chuan Neo Chong, Chairwoman of Accenture Greater China, told the South China Morning Post.
Sourced from Accenture, Vision.ee, Gartner, South China Morning Post, WARC; additional content by WARC staff