Google is absolutely in love with machine learning, and everybody even passingly familiar with the tech world is well aware of that fact. They want machine learning in all of their products, and are willing to turn a team of everyday coders into machine learning ninjas to reach that end. This, however, does not mean that they are the undisputed chart-topper in all fields of machine learning. One field that machine learning has been applied in where Google is not the top dog is object identification. Although their Google Goggles product does use AI to some extent in identifying objects that users take pictures of, a startup called Moodstocks does things a bit different, which is why Google announced on Wednesday that they've bought them up.
The French startup creates software that uses machine learning and mostly on-device computing to quickly and accurately identify random objects. Their technology consists of brand new algorithms, written just for this particular task. Google's purchase of the company reflects their increasing focus on machine learning to improve current products and churn out new ones. The arrangement calls for Moodstocks to cease its own operations and immediately transition to working on Google projects and products. They will move out of their own offices and into Google's research center in Paris, opened up in 2011, to work alongside Google's own engineers on implementing their technology into Google products.
Coincidentally, this comes as Google is rumored to be putting an image search of sorts into Google Photos, allowing users to access a Google-esque search by image functionality within their own existing photo library. Google Now On Tap also recently scored photo recognition features. Obviously, this acquisition would be helpful with things like Google's "Search Google for image" function, as well. Google's blog post did not go into much detail on any possible current or future projects that would use Moodstocks' technology and engineering talent, only saying that the acquisition is further proof of their devotion to company-wide integration of machine learning into all products. They did say that the engineers in the Paris office are working on Chrome, YouTube and Cultural Institute, which means that Moodstocks staff may be working alongside them on these projects.