What Happens When a Neural Net Colourizes Images?

We’ve already seen what happens to images when neural networks attempt to process them with Google’s DeepDream providing results varying from weird to nightmarish. This time, a neural network has been let loose with a far less potential for horror, the task? To colourize a number of black and white images with what it believes to be the correct colours based on analysis of a large number of similar images. This is the work of a team of University of California at Berkeley researchers who experimented with the neural network, with the findings being published in a paper titled Colorful Image Colorization.

The results are a mixed bag, with many of the images with simpler or more limited colour palettes seeming almost spot on, while some of the most complex images, such as the Monarch butterfly, are very impressive. Primarily, the algorithm follows simple rules that seem obvious to us, such as areas identified as “sky” will be blue, as will water or that dirt will be brown and developing from this a coloured picture. There are some clear flaws in certain images, though, with colours following outside of the “lines” in more detailed and complex images, and some difficulties with white, resulting in a rather discoloured heron.

The images were also put through a “colorization Turing Test”, where the images were able to fool human observers into believing that they were not originally monochrome images 20% of the time, which may not sound impressive, but when 50% is the expected rate was it almost impossible to distinguish, it is much more so. With neural networks already producing such results from image analysis, you can’t help but image what wonders (or horrors) they will produce next.

Oculus Rift Team Working on New Face Tracking Technology

Virtual reality technology hasn’t even hit the market yet, but that won’t stop the Oculus Rift team from developing new technology it seems. Word is that the Facebook-owned company is now looking into a way to capture and display your facial expressions in real-time.

Oculus already provides a way for users to interact with a completely different reality, but this new technology may skyrocket the realism even further by adding a key feature to making the virtual environment more lifelike.

In other words, picture two avatars inside a virtual world, each with the ability to interact and express their emotions through facial expressions. It’s as cool as it is scary, isn’t it? This may even fully immerse you in the virtual world and make you forget you are actually there.

The Oculus team is working with a team of researchers from the University of California in order to develop this new technology. They apparently came up with two designs, one involving a foam padding that covers the forehead. The latter was able to capture brow and some eye muscle movement.

The second design is a bit weird, involving a short adjustable boom attached to the headset. Both designs are able to capture data and send it to be analyzed by the software, which in turn transforms it into facial expressions. Though the technology is currently used only for research purposes, it could in theory be modified to a consumer ready device. The question is, will users be interested in taking their facial expressions to the virtual world?

Thank you Phys.org for providing us with this information

Google Pulls 200 Chrome Extensions Over Malware Fears

Google has killed nearly 200 extensions for its Chrome browser after discovering that they were injecting ads, and possibly malware, into users’ systems.

A Google study, carried out with the University of California at Berkeley, found that 192 Chrome extensions – over one-third of the total extensions available – injected deceptive ads into the browser, some of which linked to malware, potentially affecting up to 14 million users.

The study was not confined to just Chrome, finding that more than 5% of visitors to Google – across multiple browser and systems, both OS X and Windows – had at least one ad injector installed in their browser, with half of them having at least two ad injectors, and one-third of them with at least four.

Though ad injectors themselves are not barred from Chrome extensions, Google tries to restrict them as much as possible, and any extension that fulfils such a purpose must declare so as part of Chrome’s terms of service.

Source: Ars Technica