We introduce a novel approach to dynamic obstacle avoidance based on Deep Reinforcement Learning by defining a traffic type independent environment with variable complexity. Filling a gap in the current literature, we thoroughly investigate the effect of missing velocity information on an agent's performance in obstacle avoidance tasks. This is a crucial issue in practice since several sensors yield only positional information of objects or vehicles. We evaluate frequently-applied approaches in scenarios of partial observability, namely the incorporation of recurrency in the deep neural networks and simple frame-stacking. For our analysis, we rely on state-of-the-art model-free deep RL algorithms. The lack of velocity information is found to significantly impact the performance of an agent. Both approaches - recurrency and frame-stacking - cannot consistently replace missing velocity information in the observation space. However, in simplified scenarios, they can significantly boost performance and stabilize the overall training procedure.
%0 Journal Article
%1 Hart2021-mr
%A Hart, Fabian
%A Waltz, Martin
%A Okhrin, Ostap
%D 2021
%I arXiv
%K
%T Missing velocity in dynamic obstacle avoidance based on Deep Reinforcement Learning
%X We introduce a novel approach to dynamic obstacle avoidance based on Deep Reinforcement Learning by defining a traffic type independent environment with variable complexity. Filling a gap in the current literature, we thoroughly investigate the effect of missing velocity information on an agent's performance in obstacle avoidance tasks. This is a crucial issue in practice since several sensors yield only positional information of objects or vehicles. We evaluate frequently-applied approaches in scenarios of partial observability, namely the incorporation of recurrency in the deep neural networks and simple frame-stacking. For our analysis, we rely on state-of-the-art model-free deep RL algorithms. The lack of velocity information is found to significantly impact the performance of an agent. Both approaches - recurrency and frame-stacking - cannot consistently replace missing velocity information in the observation space. However, in simplified scenarios, they can significantly boost performance and stabilize the overall training procedure.
@article{Hart2021-mr,
abstract = {We introduce a novel approach to dynamic obstacle avoidance based on Deep Reinforcement Learning by defining a traffic type independent environment with variable complexity. Filling a gap in the current literature, we thoroughly investigate the effect of missing velocity information on an agent's performance in obstacle avoidance tasks. This is a crucial issue in practice since several sensors yield only positional information of objects or vehicles. We evaluate frequently-applied approaches in scenarios of partial observability, namely the incorporation of recurrency in the deep neural networks and simple frame-stacking. For our analysis, we rely on state-of-the-art model-free deep RL algorithms. The lack of velocity information is found to significantly impact the performance of an agent. Both approaches - recurrency and frame-stacking - cannot consistently replace missing velocity information in the observation space. However, in simplified scenarios, they can significantly boost performance and stabilize the overall training procedure.},
added-at = {2024-09-10T11:56:37.000+0200},
author = {Hart, Fabian and Waltz, Martin and Okhrin, Ostap},
biburl = {https://puma.scadsai.uni-leipzig.de/bibtex/2951c2fe6cc1ef3e28f8391f399f6454c/scadsfct},
interhash = {870622ce0146b65d832f762b4e290a19},
intrahash = {951c2fe6cc1ef3e28f8391f399f6454c},
keywords = {},
publisher = {arXiv},
timestamp = {2024-09-10T15:15:57.000+0200},
title = {Missing velocity in dynamic obstacle avoidance based on Deep Reinforcement Learning},
year = 2021
}
