Can machines dream? New images released by Google have given us a potential answer.
The images range from beautiful to rather bizarre and were created by the image recognition software that Google use. Over time, the system has been “taught” to recognize buildings, animals and much more in the photographs.
In plain speak, the system works by being fed a picture, processing it, then modifying it to emphasise the parts it recognizes. Eventually, the system loops and loops until the image is unrecognizable. At a low level, the neural network might be tasked merely to detect the edges on an image. In that case, the picture becomes painterly, an effect that will be instantly familiar to anyone who has experience playing about with photoshop filters.
But if the neural network is tasked with finding a more complex feature – such as animals – in an image, it ends up generating a much more disturbing hallucination.
The image recognition software used to generate these has already made it into consumer products. Google’s new photo service, Google Photos, features the option to search images with text: entering “dog”, for instance, will pull out every image Google can find which has a dog in it.
So there you have it: Androids don’t just dream of electric sheep; they also dream of mesmerising, multicoloured landscapes.
Thank you to The Guardian for providing us with this information.
Electronic Arts (EA) announced today that its games were played for over 11 billion hours…
Steam's annual end-of-year recap, Steam Replay, provides fascinating insights into gamer habits by comparing individual…
GSC GameWorld released a major title update for STALKER 2 this seeking, bringing the game…
Without any formal announcement, Intel appears to have revealed its new Core 200H series processors…
Ubisoft is not having the best of times, but despite recent flops, the company still…
If you haven’t started playing STALKER 2: Heart of Chornobyl yet, now might be the…