Computers are larger dimensions –
Now computers have the power to drive a car, can even defeat champions in chess and many more. The revolution in this intelligence comes largely just from the power of the artificial neural network. The design of this is basically inspired by the connection layer of neurons in the mammalian visual cortex. The original story is actually reprinted with permission from quanta magazine, which is an editorially independent publication of the Simons Foundation. The main mission of this is to improve the understanding of public science by covering research development and trends in mathematics, physical as well as biological sciences. If you applied this without any built-in flat geometry, it would not work that well.
Neural networks –
The researchers have actually delivered a new theoretical framework, which is actually to build a neural network that can learn patterns on many types of geometric surfaces. The equivalent convolution neural networks of caliber or CNN caliber is actually developed at the University of Amsterdam and Qualcomm AI research by taco Cohen, max welling, Berkey kicanaoglu. This can detect patterns of 2D arrays of pixels as well as in sphere and asymmetrically curved objects. Welling said on this that “This framework is a fairly definitive answer to this problem of deep learning on curved surfaces.”
How to find a solution to anything? The researcher’s solution is to make a deep learning work beyond flatlands and should have a deep connection with physics. Albert Einstein’s general theory of relativity and the standard model of particle physics, which basically exhibits a property of which is called a ” caliber equivalence.” Actually, the measurements that have made on these different circumstances and indicators are that it should be convertible with each other in a way that it can preserve the underlying relationship between the things.