Conditional Affordance Learning for Driving in Urban Environments

Axel Sauer, Nikolay Savinov, Andreas Geiger
Proceedings of The 2nd Conference on Robot Learning, PMLR 87:237-252, 2018.

Abstract

Most existing approaches to autonomous driving fall into one of two categories: modular pipelines, that build an extensive model of the environment, and imitation learning approaches, that map images directly to control outputs. A recently proposed third paradigm, direct perception, aims to combine the advantages of both by using a neural network to learn appropriate low-dimensional intermediate representations. However, existing direct perception approaches are restricted to simple highway situations, lacking the ability to navigate intersections, stop at traffic lights or respect speed limits. In this work, we propose a direct perception approach which maps video input to intermediate representations suitable for autonomous navigation in complex urban environments given high-level directional inputs. Compared to state-of-the-art reinforcement and conditional imitation learning approaches, we achieve an improvement of up to 68 % in goal-directed navigation on the challenging CARLA simulation benchmark. In addition, our approach is the first to handle traffic lights and speed signs by using image-level labels only, as well as smooth car-following, resulting in a significant reduction of traffic accidents in simulation.

Cite this Paper

BibTeX
@InProceedings{pmlr-v87-sauer18a, title = {Conditional Affordance Learning for Driving in Urban Environments}, author = {Sauer, Axel and Savinov, Nikolay and Geiger, Andreas}, booktitle = {Proceedings of The 2nd Conference on Robot Learning}, pages = {237--252}, year = {2018}, editor = {Billard, Aude and Dragan, Anca and Peters, Jan and Morimoto, Jun}, volume = {87}, series = {Proceedings of Machine Learning Research}, month = {29--31 Oct}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v87/sauer18a/sauer18a.pdf}, url = {https://proceedings.mlr.press/v87/sauer18a.html}, abstract = {Most existing approaches to autonomous driving fall into one of two categories: modular pipelines, that build an extensive model of the environment, and imitation learning approaches, that map images directly to control outputs. A recently proposed third paradigm, direct perception, aims to combine the advantages of both by using a neural network to learn appropriate low-dimensional intermediate representations. However, existing direct perception approaches are restricted to simple highway situations, lacking the ability to navigate intersections, stop at traffic lights or respect speed limits. In this work, we propose a direct perception approach which maps video input to intermediate representations suitable for autonomous navigation in complex urban environments given high-level directional inputs. Compared to state-of-the-art reinforcement and conditional imitation learning approaches, we achieve an improvement of up to 68 % in goal-directed navigation on the challenging CARLA simulation benchmark. In addition, our approach is the first to handle traffic lights and speed signs by using image-level labels only, as well as smooth car-following, resulting in a significant reduction of traffic accidents in simulation. } }
Endnote
%0 Conference Paper %T Conditional Affordance Learning for Driving in Urban Environments %A Axel Sauer %A Nikolay Savinov %A Andreas Geiger %B Proceedings of The 2nd Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2018 %E Aude Billard %E Anca Dragan %E Jan Peters %E Jun Morimoto %F pmlr-v87-sauer18a %I PMLR %P 237--252 %U https://proceedings.mlr.press/v87/sauer18a.html %V 87 %X Most existing approaches to autonomous driving fall into one of two categories: modular pipelines, that build an extensive model of the environment, and imitation learning approaches, that map images directly to control outputs. A recently proposed third paradigm, direct perception, aims to combine the advantages of both by using a neural network to learn appropriate low-dimensional intermediate representations. However, existing direct perception approaches are restricted to simple highway situations, lacking the ability to navigate intersections, stop at traffic lights or respect speed limits. In this work, we propose a direct perception approach which maps video input to intermediate representations suitable for autonomous navigation in complex urban environments given high-level directional inputs. Compared to state-of-the-art reinforcement and conditional imitation learning approaches, we achieve an improvement of up to 68 % in goal-directed navigation on the challenging CARLA simulation benchmark. In addition, our approach is the first to handle traffic lights and speed signs by using image-level labels only, as well as smooth car-following, resulting in a significant reduction of traffic accidents in simulation.
APA
Sauer, A., Savinov, N. & Geiger, A.. (2018). Conditional Affordance Learning for Driving in Urban Environments. Proceedings of The 2nd Conference on Robot Learning, in Proceedings of Machine Learning Research 87:237-252 Available from https://proceedings.mlr.press/v87/sauer18a.html.

Related Material