In the laboratories of Google are working hard to develop the best solutions for autonomous guide, the border seems more and more promising, but still has clear limits.
Of the rest, transforming a car into a smart object, or at least try to get closer to the elasticity, capacity and speed of calculation and learning of the human brain, is an unprecedented challenge.
In the laboratories dedicated to the artificial intelligence of Google, called Google Brain, are putting up the ViDAR, a technology presented at the Conference on Computer Vision and Pattern Recognition (CVPR) 2020 which is very promising.
The system ViDAR is able to recognize the movement of distant objects, by extending the depth limitations of the current sensors video.
In practice, thanks to the artificial intelligence that can estimate the camera motion with respect to the environment by enabling a greater accuracy in the scanning of the surrounding environment and predict the direction that will take the other vehicles or other road users, pedestrians included.
In this way, you go beyond the limits of today’s cameras, that speed growing are subject to a greater effect “jelly” due to the vibration, a distortion of the images that it produces a certain margin of error in communicating the input to the various servo motors that make up a system of autonomous guide.