Every View Counts: Cross-View Consistency in 3D Object Detection with Hybrid-Cylindrical-Spherical Voxelization

no code implementations • NeurIPS 2020 • Qi Chen, Lin Sun, Ernest Cheung, Alan L. Yuille

We proposed a pair of cross-view transformers to transform the feature maps into the other view and introduce cross-view consistency loss on them.

3D Object Detection Autonomous Vehicles +1

Identifying Driver Behaviors using Trajectory Features for Vehicle Navigation

no code implementations • 2 Mar 2018 • Ernest Cheung, Aniket Bera, Emily Kubin, Kurt Gray, Dinesh Manocha

We present a novel approach to automatically identify driver behaviors from vehicle trajectories and use them for safe navigation of autonomous vehicles.

Robotics

LCrowdV: Generating Labeled Videos for Simulation-based Crowd Behavior Learning

no code implementations • 29 Jun 2016 • Ernest Cheung, Tsan Kwong Wong, Aniket Bera, Xiaogang Wang, Dinesh Manocha

We present a novel procedural framework to generate an arbitrary number of labeled crowd videos (LCrowdV).

General Classification Pedestrian Detection +1