Image recognition technology is slowly but surely going mainstream. Earlier this year, Amazon added a feature to its mobile app that promised to make shopping on their platform even easier than it was before. The feature, called Flow, lets users identify products simply by snapping a photo of it. Users can then purchase the product instantly on Amazon – remarkably simple and efficient.
The thing about image recognition as a feature is that it needs to be lightning quick. Human beings have attention spans that are progressively getting shorter, and extended loading times are a surefire way of losing a potential customer on the spot. For this reason, the team at Taiwan-based visual search service Viscovery made speed the top priority when developing their image recognition technology.
Connie Huang, business development manager at Viscovery, claims that they have the best recognition speed in the market right now. How good is it, you ask? Here’s a video that demonstrates how fast the process of recognition is:
As can be seen, Viscovery is markedly faster than other apps in recognizing the products from the image taken.
How was the team able to achieve this? Offline image recognition is the answer. According to Huang, images that are taken do not need to be uploaded to the main server to be recognized – the recognition occurs within the mobile app itself. As a result, recognition of the image is extremely quick.
In addition to speed, taking image recognition offline could come in especially handy within Asia, where internet penetration rates are still fairly low, though certainly on the way up. Viscovery is able to help merchants reach out to a far larger market segment than they would have been able to with conventional cloud-based image recognition technology.
Making it more accessible
The company started off working on Chinese character translation using visual search. Like many fledging startups, they were initially unaware of the broad applicability of the technology they had on hand.
“Eventually Viscovery found its identity and strength in developing visual search technology for mobile devices. The proliferation of mobile devices such as smartphones created ample business opportunities for visual search technologies to be used in everyday life,” Huang explains. “Our research and development team then zoomed in to focus on the development of solutions such as cloud-based image recognition and offline search using wearable technology.”
From there, they began to identify several industries which would benefit greatly from instant information provision:
We have since signed up major clients in the consumer health product segment such as Watson and Amway, as well as book retailers and cellphone operators in Asia. On the consumer end, the convenience of using smartphones to identify, learn, and interact both online and offline really improves the consumer e-commerce experience tremendously. We believe the potential for our products is immense.
Viscovery has entered a couple of startup events and has seen some interest in the technology thus far. “Our future plan involves efforts to seek well-placed partners to scale our products in Asia, and eventually globally,” Huang says. “We are well underway on that plan, and it is great to see the excitement from our initial focus markets through events such as GMIC in China and Belaunch in Korea.”
The team was chosen as one of the top 10 startups at GMIC. They have received seed funding from Pinehurst Advisors, and are currently looking to raise a round of series A funding to accelerate growth both in Asia and beyond.Editing by Steven Millward