• Corpus ID: 214151340

Scalability in Perception for Autonomous Driving: An Open Dataset Benchmark

  title={Scalability in Perception for Autonomous Driving: An Open Dataset Benchmark},
  author={Pei Sun and Henrik Kretzschmar and Xerxes Dotiwalla and Aurelien Chouard and Vijaysai Patnaik and Paul Tsui and James Guo and Yin Zhou and Yuning Chai and Benjamin Caine and Vijay Vasudevan and Wei Han and Jiquan Ngiam and Hang Zhao and Aleksei Timofeev and Scott M. Ettinger and Maxim Krivokon and Amy Gao and Aditya Joshi and Yuzhang and Jonathon Shlens and Z. Chen and Dragomir Anguelov},
Simulating the Autonomous Future: A Look at Virtual Vehicle Environments and How to Validate Simulation Using Public Data Sets
A proposed framework for constructing, parameterizing, and validating a virtual vehicle environment using an existing AV data set is examined, including an overview of several open source and commercially available simulation tools for scene and scenario creation.
Point Density-Aware Voxels for LiDAR 3D Object Detection
Point Density-Aware Voxel network (PDV), is an end-to-end two stage LiDAR 3D object detection architecture that outperforms all state-of-the-art methods on the Waymo Open Dataset and achieves competitive results on the KITTI dataset.
Center-based 3D Object Detection and Tracking
The framework, CenterPoint, first detects centers of objects using a keypoint detector and regresses to other attributes, including 3D size, 3D orientation, and velocity, and refines these estimates using additional point features on the object.
DV-Det: Efficient 3D Point Cloud Object Detection with Dynamic Voxelization
This work proposes dynamic voxelization, a method that voxellizes points at local scale on-the-fly to preserve the point cloud geometry with 3D voxels, and waive the dependence on expensive MLPs to learn from point coordinates.
Score refinement for confidence-based 3D multi-object tracking
This work focuses on a neglected part of the tracking system: score refinement and tracklet termination, and shows that manipulating the scores depending on time consistency while terminating the tracklets depending on the tracklet score improves tracking results.
Unsupervised Object Detection with LiDAR Clues
This paper exploits LiDAR clues to aid unsupervised object detection by exploiting the 3D scene structure, and identifies another major issue, seldom noticed by the community, that the long-tailed and open-ended (sub-)category distribution should be accommodated.
Seeing Through Fog Without Seeing Fog: Deep Multimodal Sensor Fusion in Unseen Adverse Weather
A deep fusion network for robust fusion without a large corpus of labeled training data covering all asymmetric distortions is presented, and a single-shot model that adaptively fuses features, driven by measurement entropy is proposed.
HOTA: A Higher Order Metric for Evaluating Multi-object Tracking
This work presents a novel MOT evaluation metric, higher order tracking accuracy (HOTA), which explicitly balances the effect of performing accurate detection, association and localization into a single unified metric for comparing trackers.
3D Object Detection with Enriched Point Cloud
  • Dazhi Cheng
  • Computer Science, Environmental Science
  • 2020
This thesis proposes a multi-view segmentationdetection fusion framework that enhances metric-preserving yet sparse voxel feature learning by dense observations from perspective view and significantly boosts performance of state-of-the-art detectors by introducing multiple points per beam.
ECP2.5D - Person Localization in Traffic Scenes
The recently released EuroCity Persons detection dataset is enhanced, a large and diverse automotive dataset covering pedestrians and riders, and an automatic 3D lifting procedure is introduced by using additional LiDAR distance measurements to augment a large part of the reasonable subset of 2D box annotations with their corresponding 3D point positions.