PLOP: Probabilistic Polynomial Objects trajectory Prediction for autonomous driving

Thibault Buhet, Emilie Wirbel, Andrei Bursuc, Xavier Perrotton
Proceedings of the 2020 Conference on Robot Learning, PMLR 155:329-338, 2021.

Abstract

To navigate safely in urban environments, an autonomous vehicle (ego vehicle) must understand and anticipate its surroundings, in particular the behavior and intents of other road users (neighbors). Most of the times, multiple decision choices are acceptable for all road users (e.g., turn right or left, or different ways of avoiding an obstacle), leading to a highly uncertain and multi-modal decision space. We focus here on predicting multiple feasible future trajectories for both ego vehicle and neighbors through a probabilistic framework. We rely on a conditional imitation learning algorithm, conditioned by a navigation command for the ego vehicle (e.g., “turn right”). Our model processes ego vehicle front-facing camera images and bird-eye view grid, computed from Lidar point clouds, with detections of past and present objects, in order to generate multiple trajectories for both ego vehicle and its neighbors. Our approach is computationally efficient and relies only on on-board sensors. We evaluate our method offline on the publicly available dataset nuScenes, achieving state-of-the-art performance, investigate the impact of our architecture choices on online simulated experiments and show preliminary insights for real vehicle control.

Cite this Paper


BibTeX
@InProceedings{pmlr-v155-buhet21a, title = {PLOP: Probabilistic Polynomial Objects trajectory Prediction for autonomous driving}, author = {Buhet, Thibault and Wirbel, Emilie and Bursuc, Andrei and Perrotton, Xavier}, booktitle = {Proceedings of the 2020 Conference on Robot Learning}, pages = {329--338}, year = {2021}, editor = {Kober, Jens and Ramos, Fabio and Tomlin, Claire}, volume = {155}, series = {Proceedings of Machine Learning Research}, month = {16--18 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v155/buhet21a/buhet21a.pdf}, url = {https://proceedings.mlr.press/v155/buhet21a.html}, abstract = {To navigate safely in urban environments, an autonomous vehicle (ego vehicle) must understand and anticipate its surroundings, in particular the behavior and intents of other road users (neighbors). Most of the times, multiple decision choices are acceptable for all road users (e.g., turn right or left, or different ways of avoiding an obstacle), leading to a highly uncertain and multi-modal decision space. We focus here on predicting multiple feasible future trajectories for both ego vehicle and neighbors through a probabilistic framework. We rely on a conditional imitation learning algorithm, conditioned by a navigation command for the ego vehicle (e.g., “turn right”). Our model processes ego vehicle front-facing camera images and bird-eye view grid, computed from Lidar point clouds, with detections of past and present objects, in order to generate multiple trajectories for both ego vehicle and its neighbors. Our approach is computationally efficient and relies only on on-board sensors. We evaluate our method offline on the publicly available dataset nuScenes, achieving state-of-the-art performance, investigate the impact of our architecture choices on online simulated experiments and show preliminary insights for real vehicle control.} }
Endnote
%0 Conference Paper %T PLOP: Probabilistic Polynomial Objects trajectory Prediction for autonomous driving %A Thibault Buhet %A Emilie Wirbel %A Andrei Bursuc %A Xavier Perrotton %B Proceedings of the 2020 Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2021 %E Jens Kober %E Fabio Ramos %E Claire Tomlin %F pmlr-v155-buhet21a %I PMLR %P 329--338 %U https://proceedings.mlr.press/v155/buhet21a.html %V 155 %X To navigate safely in urban environments, an autonomous vehicle (ego vehicle) must understand and anticipate its surroundings, in particular the behavior and intents of other road users (neighbors). Most of the times, multiple decision choices are acceptable for all road users (e.g., turn right or left, or different ways of avoiding an obstacle), leading to a highly uncertain and multi-modal decision space. We focus here on predicting multiple feasible future trajectories for both ego vehicle and neighbors through a probabilistic framework. We rely on a conditional imitation learning algorithm, conditioned by a navigation command for the ego vehicle (e.g., “turn right”). Our model processes ego vehicle front-facing camera images and bird-eye view grid, computed from Lidar point clouds, with detections of past and present objects, in order to generate multiple trajectories for both ego vehicle and its neighbors. Our approach is computationally efficient and relies only on on-board sensors. We evaluate our method offline on the publicly available dataset nuScenes, achieving state-of-the-art performance, investigate the impact of our architecture choices on online simulated experiments and show preliminary insights for real vehicle control.
APA
Buhet, T., Wirbel, E., Bursuc, A. & Perrotton, X.. (2021). PLOP: Probabilistic Polynomial Objects trajectory Prediction for autonomous driving. Proceedings of the 2020 Conference on Robot Learning, in Proceedings of Machine Learning Research 155:329-338 Available from https://proceedings.mlr.press/v155/buhet21a.html.

Related Material