Pillar-Based Object Detection for Autonomous Driving
Name
2007.10323.pdf
Description
Submitted version
Size
2.06 MB
Format
Adobe PDF
Checksum (MD5)
e2a349c1b248f2f3416c18f38092668c
Author(s)
Wang, Y
Fathi, A
Kundu, A
Ross, DA
Pantofaru, C
Funkhouser, T
Solomon, J
Date Issued
July 2020
Journal
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher
Springer International Publishing
Citation
Wang, Y, Fathi, A, Kundu, A, Ross, DA, Pantofaru, C et al. 2020. "Pillar-Based Object Detection for Autonomous Driving." Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 12367 LNCS.
Version
Original manuscript
Abstract
© 2020, Springer Nature Switzerland AG. We present a simple and flexible object detection framework optimized for autonomous driving. Building on the observation that point clouds in this application are extremely sparse, we propose a practical pillar-based approach to fix the imbalance issue caused by anchors. In particular, our algorithm incorporates a cylindrical projection into multi-view feature learning, predicts bounding box parameters per pillar rather than per point or per anchor, and includes an aligned pillar-to-point projection module to improve the final prediction. Our anchor-free approach avoids hyperparameter search associated with past methods, simplifying 3D object detection while significantly improving upon state-of-the-art.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1007/978-3-030-58542-6_2
