5189346173_3890eb45e3_amazon

Is Amazon serious about visual recognition ?

Amazon is the latest big company to throw itself into the visual content recognition market.  Armed with the late 2015 acquisition of Orbeus, makers of ReKognition and Photo Time, it just launched its own APi service, also called ReKognition. After giants like Google, IBM, Microsoft, and smaller companies like Imagga or Clarifai, it is the latest, and maybe most formidable, offering in this space

At first glance, nothing really new here, if compared to other similar services. In fact, it seems that Amazon is done little to enhance the services already offered by Orbeus. Everything from content recognition ( including dog breeds) to face recognition were already available.  But like anything in the deep learning space, most of the advantages can really only be seen under the hood.  Amazon has a huge volume of classified  data which can be used to better train its algorithm and has most certainly use some of it.

Amazon face recognition with attributes
Amazon face recognition with attributes

What really puts Amazon as a challenging competitor is its integration with its cloud service. Already above 1 Million active users, it can certainly count on converting  a substantial amount to it visual recognition services, allured by the ease of management.   On the flip side, it might be such of appeal for those using other cloud services or simply using other solution. In fact, much like Google or Microsoft, Amazon Rekognition is more of an add-on than an independent service.

Priced under Google Vision but higher than Imagga, it seems to aim primarily at the higher end of the market, those with massive amount of data to process, rather than the up and coming startup. It doesn’t offer simple custom training like Clarifai or Imagga does, which will probably leave a lot of more specialized users out of their reach. It is currently available in 3 regions, the US ( East, West) and Ireland.

Amazon visual recognition can recognize breeds
Amazon visual recognition can recognize breeds

The biggest surprise, certainly, is the complete lack of product/logo recognition offering. Considering Amazon’s core business, it is certainly a solution they could have easily offered. However, rather than a technical issue, it is probably the concern of helping a competitor emerge that probably  fueled their decision not to have included it.

Apparently, Amazon chose the “us too” route rather than becoming a leading innovator. Nevertheless, Amazon can now be counted into the closed club of visual content recognition mighty developers and they will certainly grab a good chunk of the marketplace, even if they are late in the game, if only because of its convenience.

Photo by Martyn Wright

Author: Paul Melcher

Paul Melcher is the founder of Kaptur. He is an entrepreneur, advisor, consultant with a strong background in licensing, copyright, sales, marketing and technology with more than 20 years experience in developing world-renowned photo based companies with two successful exits. Named one of the “100 most influential people in photography” by American Photo magazine.

Share This!Share on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Share on RedditBuffer this pageEmail this to someone

Leave a Reply