One of the critical pieces of the self-driving puzzle is understanding the surroundings of a self-driving vehicle (SDV) and predicting how these surroundings will change in the near future. To address this task we propose MultiXNet, an end-to-end approach for detection and motion prediction based directly on lidar sensor data. This approach builds on prior work by handling multiple classes of traffic actors, adding a jointly trained second-stage trajectory refinement step, and producing a multimodal probability distribution over future actor motion that includes both multiple discrete traffic behaviors and calibrated continuous position uncertainties. The method was evaluated on large-scale, real-world data collected by a fleet of SDVs in several cities, with the results indicating that it outperforms existing state-of-the-art approaches.
@article{arxiv.2006.02000,
title = {MultiXNet: Multiclass Multistage Multimodal Motion Prediction},
author = {Nemanja Djuric and Henggang Cui and Zhaoen Su and Shangxuan Wu and Huahua Wang and Fang-Chieh Chou and Luisa San Martin and Song Feng and Rui Hu and Yang Xu and Alyssa Dayan and Sidney Zhang and Brian C. Becker and Gregory P. Meyer and Carlos Vallespi-Gonzalez and Carl K. Wellington},
journal= {arXiv preprint arXiv:2006.02000},
year = {2021}
}
Comments
Accepted for publication at IEEE Intelligent Vehicles Symposium (IV) 2021