This story draft by @escholar has not been reviewed by an editor, YET.

Deep Stochastic Kinematic Models for Probabilistic Motion Forecasting in Traffic: Appendix

EScholar: Electronic Academic Papers for Scholars HackerNoon profile picture
0-item

Table of Links

Abstract and I. Introduction

II. Related Work

III. Kinematics of Traffic Agents

IV. Methodology

V. Results

VI. Discussion and Conclusion, and References

VII. Appendix

VII. APPENDIX

A. Full Expansion of Formulation 3



B. Additional Results By Class


In the paper, we present results on vehicles since we use kinematic models based on vehicles as priors. Here, we present the full results per-class for each experiment in Tables IV, V, VI, and VII. The results reported in the paper are starred (*), which are re-iterated below for full context.


TABLE IV: Per-class results for performance on 100% of the Waymo Dataset.


TABLE V: Per-class results for performance on 1% of the Waymo Dataset.


TABLE VI: Per-class performance degradation results with perturbed evaluation for models trained on 100% of the Waymo Dataset.


TABLE VII: Per-class performance degradation results with perturbed evaluation for models trained on 1% of the Waymo Dataset.


C. Experiment Hyperparameters


TABLE VIII: Model Architecture Hyperparameters


Authors:

(1) Laura Zheng, Department of Computer Science, University of Maryland at College Park, MD, U.S.A ([email protected]);

(2) Sanghyun Son, Department of Computer Science, University of Maryland at College Park, MD, U.S.A ([email protected]);

(3) Jing Liang, Department of Computer Science, University of Maryland at College Park, MD, U.S.A ([email protected]);

(4) Xijun Wang, Department of Computer Science, University of Maryland at College Park, MD, U.S.A ([email protected]);

(5) Brian Clipp, Kitware ([email protected]);

(6) Ming C. Lin, Department of Computer Science, University of Maryland at College Park, MD, U.S.A ([email protected]).


This paper is available on arxiv under ATTRIBUTION-NONCOMMERCIAL-NODERIVS 4.0 INTERNATIONAL license.


L O A D I N G
. . . comments & more!

About Author

EScholar: Electronic Academic Papers for Scholars HackerNoon profile picture
EScholar: Electronic Academic Papers for Scholars@escholar
We publish the best academic work (that's too often lost to peer reviews & the TA's desk) to the global tech community

Topics

Around The Web...

Trending Topics

blockchaincryptocurrencyhackernoon-top-storyprogrammingsoftware-developmenttechnologystartuphackernoon-booksBitcoinbooks