Apple Introduces "VoxelNet" In New Research Paper

Yolanda Curtis
November 22, 2017

Research by Apple computer scientists on how self-driving cars can better spot cyclists and pedestrians while using fewer sensors has been posted online, in what appears to be the company's first publicly disclosed paper on autonomous vehicles.

The research paper, submitted last week to the journal arXiv, was written by Apple scientists Yin Zhou and Oncel Tuzel. Not the real Apple Car. "Future work includes extending VoxelNet for joint LiDAR and image based end-to-end 3D detection to further improve detection and localization accuracy".

Denser point clouds offer a clearer, more accurate picture of an object, but Apple's researchers say their new method, which they call VoxelNet, makes even sparse point clouds useful for object detection.

With the new software, engineers said they produced "highly encouraging results" in terms of identifying pedestrians and cyclists using only LiDAR data, which gathers information about surroundings using a pulsed laser, not cameras. Apple's researchers argue a LiDAR and camera-based solution could be hard to deploy in many situations and it could be "more sensitive to sensor failure". We know Apple's working on this because it's had to admit as much in order to secure a self-driving test permit from the California Department of Motor Vehicles, and because its test auto has been spotted in and around time. The results come from the use of simulators, not Apple's on-road efforts.

The paper seems to confirm rumors that Apple had nixed its plans for developing an actual self-driving vehicle to compete with Tesla, and pivoted instead to working on autonomous software platforms. This move does not come as a surprise, Bryant Walker Smith, assistant professor of law at the University of SC and expert in the autonomous vehicles space, told TechRepublic amid earlier Apple self-driving auto rumors.

While the scientific part of the paper is interesting, the most surprising part about it is that it exists at all. "And we sort of see it as the mother of all AI projects".

Other reports by iNewsToday