Use lasers to identify cyclists and pedestrians

Apple has revealed details of its self-driving car system for the first time. The system USES a new combination of software and laser detection to identify pedestrians and cyclists.

While apple has never hidden its ambitions for self-driving cars, little has been known about the secretive project until now. Tim cook, apple’s chief executive, said: Cook calls self-driving cars “the mother of all ai projects,” but apple rarely reveals much about them. But now, two of the company’s computer scientists have published a paper online that reveals how self-driving cars work. According to the paper, apple’s self-driving cars use lasaer pointer sensors in new software called VoxelNet to better identify cyclists and pedestrians.

Yin Zhou and Oncel Tuzel published their seminal paper in the independent online journal arXiv on November 17. Because apple is notoriously secretive about new products, it has long been a liability for ai and machine researchers. In this paper, a new VoxelNet software is proposed to help computer detect 3d objects.

Apple, however, has not commented on the paper. Developers are accustomed to freely sharing their work with peers in other organizations. To that end, apple in July created the apple machine self-learning journal for its research team. The developer’s work log rarely appears outside the apple machine self-learning journal, where no research on self-driving cars has been published so far.

Self-driving cars often use a combination of ordinary two-dimensional cameras and depth-sensing lidar to identify their surroundings. While these combinations provide depth information, their low resolution makes it difficult to detect small objects at great distances without a real-time link to a regular camera.

However, according to apple researchers, the use of the new software has yielded promising results in identifying pedestrians and cyclists. In addition, using the software is an advantage over other methods that only use lidar to detect three-dimensional objects. But the results are from computer simulations, not road tests.