Autonomous Driving at Intersections: A Critical-Turning-Point Approach for Left Turns

Shu, Keqi; Yu, Hongkai; Chen, X.; Chen, L.; Wang, Q.; Li, L.; Cao, Dongpu

doi:10.48550/arxiv.2003.02409

Cited by 9 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We observe that our system performs slightly better on right turns over left and hypothesize that these turns are somewhat easier as the car can simply follow the right boundary of the road to complete right turns, but needs to perceive and attend to a more distant opposite boundary during a left turn. This observation is also consistent with previous literature [51].…”

Section: Intersection Number Directionsupporting

confidence: 94%

Efficient and Robust LiDAR-Based End-to-End Navigation

Liu

Amini

Zhu

et al. 2021

2021 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

Deep learning has been used to demonstrate endto-end neural network learning for autonomous vehicle control from raw sensory input. While LiDAR sensors provide reliably accurate information, existing end-to-end driving solutions are mainly based on cameras since processing 3D data requires a large memory footprint and computation cost. On the other hand, increasing the robustness of these systems is also critical; however, even estimating the model's uncertainty is very challenging due to the cost of sampling-based methods. In this paper, we present an efficient and robust LiDAR-based end-to-end navigation framework. We first introduce Fast-LiDARNet that is based on sparse convolution kernel optimization and hardwareaware model design. We then propose Hybrid Evidential Fusion that directly estimates the uncertainty of the prediction from only a single forward pass and then fuses the control predictions intelligently. We evaluate our system on a full-scale vehicle and demonstrate lane-stable as well as navigation capabilities. In the presence of out-of-distribution events (e.g., sensor failures), our system significantly improves robustness and reduces the number of takeovers in the real world.

show abstract

Section: Intersection Number Directionsupporting

confidence: 94%

Efficient and Robust LiDAR-Based End-to-End Navigation

Liu

Amini

Zhu

et al. 2021

2021 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

show abstract

“…Future work will focus on optimizing road tessellation, incorporating spatiotemporal state dependencies, and employing computer vision models for estimating the distribution of observed vehicles. These efforts will lead to developing safe-decision-making algorithms for autonomous vehicles operating in urban environments [43], [44].…”

Section: Discussionmentioning

confidence: 99%

Directional Primitives for Uncertainty-Aware Motion Estimation in Urban Environments

Senanayake¹,

Toyungyernsub²,

Wang³

et al. 2020

Preprint

View full text Add to dashboard Cite

We can use driving data collected over a long period of time to extract rich information about how vehicles behave in different areas of the roads. In this paper, we introduce the concept of directional primitives, which is a representation of prior information of road networks. Specifically, we represent the uncertainty of directions using a mixture of von Mises distributions and associated speeds using gamma distributions. These location-dependent primitives can be combined with motion information of surrounding vehicles to predict their future behavior in the form of probability distributions. Experiments conducted on highways, intersections, and roundabouts in the Carla simulator, as well as real-world urban driving datasets, indicate that primitives lead to better uncertainty-aware motion estimation.

show abstract

“…This problem can be modeled as a Markov Decision Process (MDP) where solutions can be obtained by utilizing online solvers or through optimal policy approximation using RL approaches. For instance, in [26], [27], the Adaptive Belief Tree (ABT) solver is used to solve the formulated Partially-Observable Markov Decision Process (POMDP) [28] of the decision-making problem. Authors use the Critical-Turning-Point (CTP) approach where the left turn trajectory is simply assumed as a straight line with a quarter circle curve.…”

Section: Drl For Unsignalized Intersection Traversal Problemmentioning

confidence: 99%

Self-Learned Autonomous Driving at Unsignalized Intersections: A Hierarchical Reinforced Learning Approach for Feasible Decision-Making

Al-Sharman¹,

Dempster²,

Daoud³

et al. 2022

Preprint

View full text Add to dashboard Cite

<p> Reinforcement learning-based techniques, empowered by deep-structured neural nets, have demonstrated superiority over rule-based methods in terms of making high-level behavioral decisions due to qualities related to handling large state spaces. Nonetheless, their training time, sample efficiency and the feasibility of the learnt behaviors remain key concerns. In this paper, we propose a novel hierarchical reinforcement learning-based decision-making architecture for learning left-turn policies at unsignalized intersections with feasibility guarantees. The proposed technique is comprised of two layers; a high-level learning-based behavioral planning layer which adopts soft actor-critic principles to learn high-level, non-conservative yet safe, driving behaviors, and a low-level Model Predictive Control (MPC) framework to ensure feasibility of the two-dimensional left-turn maneuver. The high-level layer generates reference signals of velocity and yaw angles for the ego vehicle taking into account safety and collision avoidance with the intersection vehicles, whereas the low-level planning layer solves an optimization problem to track these reference commands taking into account several vehicle dynamic constraints and ride comfort. We validate the proposed decision-making scheme in simulated environments and compare with other model free Reinforcement Learning (RL) baselines. The results demonstrate that the proposed integrated framework possesses better training and navigation capabilities.</p>

show abstract

Autonomous Driving at Intersections: A Critical-Turning-Point Approach for Left Turns

Cited by 9 publications

References 18 publications

Efficient and Robust LiDAR-Based End-to-End Navigation

Efficient and Robust LiDAR-Based End-to-End Navigation

Directional Primitives for Uncertainty-Aware Motion Estimation in Urban Environments

Self-Learned Autonomous Driving at Unsignalized Intersections: A Hierarchical Reinforced Learning Approach for Feasible Decision-Making

Contact Info

Product

Resources

About