Artificial intelligence in unmanned vehicles Apple will learn better to distinguish between the types of objects on the street.
Work recognition system of three-dimensional objects. From left to right: vehicle detection, pedestrian, cyclist.
According to the portal AppleInsider, the network got a document that describes a method for detecting objects in a group of 3D points using trainees neural networks. This method will be used by Apple in the heart of the LIDAR system.
LIDAR is the technology of obtaining and processing information about deleted objects using active optical systems that use reflection and dispersion of light in different environments. It is based on the principle of optical range-finder that generates a 3D map of the surrounding space. The more distance, the higher the accuracy of the data.
The new method will allow to detect and distinguish between vehicles, cyclists and pedestrians when using a much smaller number of rangefinders and sensors.
The current difficulties in the operation of systems of unmanned vehicles are that low resolution sensors and sensors difficult to distinguish small, distant objects. So you have to resort to using an ordinary camera, transmitting the image of the traffic situation in real time.
While all the experiments with the new technology developed by Apple, represent only a computer simulation, testing on the roads was not carried out.
Although Tim cook puts the development of self-driving cars at the forefront, the company does not share information about their developments in this area.
Follow the news on our Apple Telegram-channel, as well as in the app on iOS MacDigger.