Acquisition Brings Image-Recognition Firm Closer to True Computer Vision | Adweek Acquisition Brings Image-Recognition Firm Closer to True Computer Vision | Adweek
Advertisement

Pongr Picks Up Israeli Startup Sightec

Acquisition allows image-recognition company to zero-in on brand images in UCG photos

As more of the Internet’s real estate becomes populated with pictures, companies are racing to develop computer vision technology that can pick out objects in an image and convert them into data to be used for ad targeting, among other possibilities.

One such company, Pongr, has spent the past five years marketing a platform for marketers to solicit photos from people that feature a brand’s product in the frame. Underlying that platform is technology that can process photos and recognize brand photos and index them—even if a logo isn’t immediately recognizable—so that these brands photos can be matched with future photos.

Now, Pongr’s piecemeal approach is getting an adrenaline shot through the company’s acquisition of Israel-based computer vision startup Sightec.

“These guys are doing something fundamentally different [than others in the computer vision field]. There’s something really powerful here if you take their algorithmic approach,” said Pongr CEO and co-founder Jamie Thompson.

Sightec’s technology allows for sub-pixel registration which Thompson explained would let Pongr detect objects in the foreground and background of an image. “Pongr has been good to date at detecting products when they’re deliberate and promotional, but because of sub-pixel registration, we could pick up products in the background,” he said. Sightec also brings image stabilization and enhancement technology that could recognize an object in an blurry or Instagram-filtered image, making the ongoing flood of user-snapped photos less of a headache.

“The problem we’re approaching is consumer-generated photos of products greatly outnumber the volume of brand-generated photos,” Thompson said.

Basically what this means is someone could have taken a picture holding up a Pepsi can, and Pongr via Sightec would be able to identify the Coke can sitting on the table behind it. Conceivably that could let Pongr segment that user or the Web page on which that user’s photo appears as someone or somewhere Pepsi and Coke might like to advertise against.

Pongr isn’t alone in trying to illuminate the value of all those billions of photos on the Web. Google has been doing its own research and development through the crowd sourced Google Goggles product, and Curalate, Pinfluencer, Luminate and Stipple are also on the job. Facebook may pose an even greater threat should the social network put its acquisition of facial recognition firm Face.com to work indexing all of Facebook’s more than 219 billion photos and Instagram’s 5-billion-plus photos, which the photo-sharing service’s recent privacy policy changes appear to permit. Thompson said that notion of Facebook’s entry is “not crazy at all.”

“Facebook’s acquisition of Face.com ratifies our position which is that visual content is super important to what these guys do next in monetizing the assets they have,” he said.

The all-stock deal for Sightec adds five full-time computer vision scientists to Pongr’s 15-person team as well as a Tel Aviv office in addition to the company’s existing Boston and Des Moines, Iowa, locations.

Advertisement