Apple Reportedly Acquired Montreal Based Vrvana, a Mixed Reality Headset Company with Hand Tracking Technologies
Work on Apple's First Flagship Store in a Trendy Shopping District in South Korea Could be completed by Late December

Apple Research Paper Touts VoxelNet as being Superior to LiDAR Regarding Autonomous Vehicle 3D Detection Methods



Apple has decided of late to expose more of their engineering side to the world for some reason after being secretive about all current work in progress on any project. Developer discussions were pretty much restricted to their World Wide Developer Conferences. Yet Apple has recently published Machine Learning Journal issue #6 about 'Hey Siri' and Journal issue #7 discussing face detection. Now Apple's Senior AI Researcher Yin Zhou and Machine Learning Research Scientist Oncel Tuzel have published their work on how self-driving cars could better detect pedestrians and cyclists while using fewer sensors.


A Brief Summary of the Paper


"Accurate detection of objects in 3D point clouds is a central problem in many applications, such as autonomous navigation, housekeeping robots, and augmented/virtual reality. To interface a highly sparse LiDAR point cloud with a region proposal network (RPN), most existing efforts have focused on hand-crafted feature representations, for example, a bird's eye view projection. In this work, we remove the need of manual feature engineering for 3D point clouds and propose VoxelNet, a generic 3D detection network that unifies feature extraction and bounding box prediction into a single stage, end-to-end trainable deep network.


Specifically, VoxelNet divides a point cloud into equally spaced 3D voxels and transforms a group of points within each voxel into a unified feature representation through the newly introduced voxel feature encoding (VFE) layer. In this way, the point cloud is encoded as a descriptive volumetric representation, which is then connected to a RPN to generate detections. Experiments on the KITTI car detection benchmark show that VoxelNet outperforms the state-of-the-art LiDAR based 3D detection methods by a large margin."



A PDF of their Paper titled, "VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection," could be accessed here.


10.1 Bar - News

About Making Comments on our Site: Patently Apple reserves the right to post, dismiss or edit any comments. Those using abusive language or negative behavior will result in being blacklisted on Disqus.




The comments to this entry are closed.