Last time we checked in with Netra, the startup founded by MIT alums was touting image recognition technology for retail security cameras, designed to help marketers analyze customer behavior. That was at Techstars Demo Day when it was presenting last September. Things have since changed.
Netra announced on Thursday that it has closed a $1.85 million seed round and, along with it, revealed the pivot it made shortly after Techstars finished last fall. The seed round was led by NXT Ventures and Launchpad Venture Group, with participation from Mark Cuban, Zelkova Ventures, Bewind Group and "other prominent angels." Cuban has been an investor since at least 2015.
"It’s really powerful with e-commerce. This turns the world into their showroom."
Richard Lee, Netra’s CEO, told BostInno the company is now using its visual recognition and analysis software, first developed at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL), for two purposes: visual intelligence and search.
For visual intelligence, Lee said it has started a pilot contract with London-based Kantar Group, a large market research firm owned by advertising and public relations giant WPP. Basically, Kantar will use Netra’s technology to automate the process of tagging images and videos with information about what brands appear and what the demographics are of each person who appears in one — what was previously a very labor-intensive process carried out by Kantar’s employees.
The technology uses computer vision and deep learning to identify visual information that is typically unseen by market research firms, Lee said, so there’s a lot of value for advertisers and marketers to learn about what kind of people are interacting with what kind of brands on images and videos posted to various social media sites like Instagram and Twitter. Beyond brands and logos, Netra can recognize age, gender and sentiment, among other things.
"The beauty of digital is — this is a person that is female, Asian and she spends a lot of time on the beach —you can serve this more effectively," Lee said. "You can get a good sense of who they are."
Netra is also working on a visual search engine, and it is currently partnering with a major e-commerce company, which Lee declined to name, to help consumers find products that are similar to ones they take pictures of. It’s kind of like Amazon’s app where you can scan a barcode of a product at a retail store to find the same product online, except Netra’s image recognition software will let you find anything that looks similar to the furniture, apparel or footwear you take a photo of.
"It’s really powerful with e-commerce," Lee said. "This turns the world into their showroom."
In a way, it's similar to the visual shopping technology being developed by Allyke, a Wellesley-based company that raised a $1.3 million seed round earlier this year, except Allyke's technology is aimed at retailers. But there are plenty of other companies that are working on photo-tagging, including Google, Facebook and Boston-based Indico.
The seed investment, which brings Netra’s total funding to $2.5 million, will be used to accelerate product development and sales, as well as grow its current team of six people to more than 10.
