Tesla makes self-driven cars that can take you anywhere automatically. They use technology such as machine vision to see Road lines, obstacles, and other cars. But how does this work? Before we start, we are assuming that you have at least a basic understanding of neural networks.
One of the major aspects of machine vision is image recognition. Image recognition is basically achieved by using convolution neural networks polished. Neural networks obviously broken down into two major steps convolutions and pooling convolutions basically news matrices and kernels – filtered image while pooling basically emphasizes, the major points of an image both of these processes together are used to downsize data relate points of closer proximity while only maintaining and therefore emphasizing the significant characteristics of an image. This modified data can then be passed through a normal neural network to identify the image.
Tesla also uses different types of neural networks to control their cars
but how do narrow networks learn well?
Tesla has over 100,000 cars on the road today all collecting data about how people react in certain conditions situations and certain locations, this basically gives to infinite amount to train their neural network input suggests at continuation and location could be passed instant ural Network and having an output a reaction, this would be compared to the reaction that the human driver acts performed to see if the neural network made the correct decision.
A process known as back propagation can then be used to enhanced accuracy of this neural network, of course it’s very abstract, we are not talking about how neural networks work with waste activation functions by snows or threshold values. These two major parts of Tesla’s product and use alongside other aspects of the system to create an almost human-like intelligence, it can perceive with machine vision it can think with no networks and it can respond with the mechanics of the system. Tesla’s founder Elon Musk one of the leading products in our view intelligence around the world is pushing us ever closer towards a true artificial intelligence.
