Facebook is open sourcing a set of computer vision software tools that can identify both the variety and the shape of objects within photos. The tools, developed by the Facebook AI Research (FAIR) team, are called DeepMask, SharpMask, and MultiPathNet, and all three work in tandem to help break down and contextualize the contents of images. These technologies, though not in active use in consumer Facebook products right now, are similar to the software the company uses to describe photos to blind users, a feature it calls "automatic alternative text" that launched back in April. DeepMask and SharpMask are more experimental research projects and focus on what the FAIR team calls segmentation. While human beings can discern the various elements of a photograph in mere seconds, the process is much harder for computers, which perceive pixels as a series of number values corresponding to changes in color.

LONDON - FEBRUARY 03: (FILE PHOTO) In this photo illustration the facebook logo is reflected in the eye of a girl on February 3, 2008 in London, England. Social networking site'Facebook' reaches it's 5th birthday this month. It was founded in 2004 by Mark Zuckerberg from his dorm room at Harvard University with the aim to help students keep in touch over the internet. Within 24 hours 1,200 Harvard students had signed up. The site now has 150 million active users worldwide. Facebook is opening up its image-recognition artificial intelligence research to the public.

Image recognition is the process of identifying and detecting an object or a feature in a digital image or video. This concept is used in many applications like systems for factory automation, toll booth monitoring, and security surveillance. Specific image recognition applications include classifying digits using HOG features and an SVM classifier (Figure 1). Cross correlation can be used for pattern matching and target tracking as shown in Figure 2. An effective approach for image recognition includes using a technical computing environment for data analysis, visualization, and algorithm development.

Google has acquired Moodstocks, a Paris-based startup that specialises in smartphone image recognition as part of its continued efforts to boost its own artificial intelligence (AI) research, development and capabilities. Announced in a blog post on 6 July, Vincent Simonet, head of Google's research and development (R&D) centre in Paris, says the tech giant's latest purchase is proof of its commitment to the promising sector. "Many Google services use machine learning to make them simpler and more useful in everyday life such as Google Translate, Smart Reply Inbox, or the Google app," Simonet wrote in French. "We have made great strides in terms of visual recognition: Now you can search in Google Pictures such as'party' or'beach' and the application will offer you good pictures without you needing to categorise them manually. But there is still much to do in this area. And this is where Moodstocks comes in."

Google has acquired Moodstocks, a company that develops machine-learning based object recognition tech for mobile phones. The Paris-based startup will shut down its object recognition Application Programming Interface (API) after its staff joins Mountain View's Parisan R&D team, reports PC World. The purchase was made for an unknown sum, and appears like an acquihire deal. The French technology startup builds photo and object recognition software by employing deep learning techniques. The company produced a visual search API and an Android app that could identify certain kinds of objects.

Moodstocks, a Parisian startup that develops image recognition tools for smartphones, is joining Google. The companies announced the acquisition today, sans financials. Around since 2008, Moodstocks hasn't had considerable traction. But the company has tech and engineers working on machine learning, something Google cannot get enough of as it competes with rivals like Apple and Facebook for talent. And Moodstocks' core service -- "to give eyes to machines by turning cameras into smart sensors," as its parting note described -- fits with Google's vision for image search and augmented reality, where a phone (or something else) knows your physical surroundings. Also, the acquisition price may have been low thanks to wobbling global markets.

Two weeks after Twitter acquired Magic Pony to advance its machine learning smarts for improving users' experience of photos and videos on its platform, Google is following suit. Today, the maker of Android and search giant announced that it has acquired Moodstocks, a startup based out of Paris that develops machine-learning based image recognition technology for smartphones whose APIs for developers have been described as "Shazam for images." Moodstocks' API and SDK will be discontinued "soon", according to an announcement on the company's homepage. "Our focus will be to build great image recognition tools within Google, but rest assured that current paying Moodstocks customers will be able to use it until the end of their subscription," the company noted. Terms of the deal were not disclosed and it's not clear how much Moodstocks had raised: CrunchBase doesn't note any VC money, although when we first wrote about the company back in 2010 we noted that it had raised 500,000 in seed funding from European investors.

The world is quietly being reshaped by machine learning. We no longer need to teach computers how to perform complex tasks like image recognition or text translation: instead, we build systems that let them learn how to do it themselves. "It's not magic," says Greg Corrado, a senior research scientist at Google. The most powerful form of machine learning being used today, called "deep learning", builds a complex mathematical structure called a neural network based on vast quantities of data. Designed to be analogous to how a human brain works, neural networks themselves were first described in the 1930s.

In the last couple of years, magic started happening in AI. Techniques started working, or started working much better, and new techniques have appeared, especially around machine learning ('ML'), and when those were applied to some long-standing and important use cases we started getting dramatically better results. For example, the error rates for image recognition, speech recognition and natural language processing have collapsed to close to human rates, at least on some measurements. So you can say to your phone: 'show me pictures of my dog at the beach' and a speech recognition system turns the audio into text, natural language processing takes the text, works out that this is a photo query and hands it off to your photo app, and your photo app, which has used ML systems to tag your photos with'dog' and'beach', runs a database query and shows you the tagged images. There are really two things going on here - you're using voice to fill in a dialogue box for a query, and that dialogue box can run queries that might not have been possible before.

Based on the agreement, both parties will work together on new technology innovation, enhanced user experience and developing a new e-commerce platform for enterprise-level products. Together, the companies are dedicated to driving the next generation of e-commerce artificial intelligence solutions, and conducting associated research and development for high-speed interconnect products. A key technology that JD.com has developed is JD Camera, an application for image recognition and similar image search in mobile terminals. JD Camera facilitates ease-of-shopping for users by allowing customers to quickly and easily search for their favorites products with just a photo rather than detailed language descriptions. "In the future, with the help of the Joint Lab, Camera will be enhanced from general photo-based searches to more advanced imaged-based searches that will allow users to view, select and purchase from suggested recommendations with an advanced image match algorithm for such items as clothing, make-up, furniture, etc.," said Weng Zhi, vice president of technology, JD.com.