Integrating end-to-end learned steering into probabilistic autonomous driving

Hubschneider, Christian; Bauer, André; Doll, Jens; Weber, Michael; Klemm, Sebastian; Klunker, Florian; Zöllner, J. Marius

doi:10.1109/itsc.2017.8317873

Cited by 30 publications

(10 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The reinforcement learning directly used the sensor input; the other two tasks were carried out in the feature space, calculating control signals according to bias to lane center and positions of other vehicles. Hubschneider et al [100] revised the trajectory of end-to-end learning with rule-based methods according to obstacle positions in the feature space. It can be found that the driving space is considered independently in both learning-based and rule-based components of the combined decision.…”

Section: Combining Learning-based and Rule-based Decision Methodsmentioning

confidence: 99%

Driving Space for Autonomous Vehicles

et al. 2019

View full text Add to dashboard Cite

Driving space for autonomous vehicles (AVs) is a simplified representation of real driving environments that helps facilitate driving decision processes. Existing literatures present numerous methods for constructing driving spaces, which is a fundamental step in AV development. This study reviews the existing researches to gain a more systematic understanding of driving space and focuses on two questions: how to reconstruct the driving environment, and how to make driving decisions within the constructed driving space. Furthermore, the advantages and disadvantages of different types of driving space are analyzed. The study provides further understanding of the relationship between perception and decision-making and gives insight into direction of future research on driving space of AVs.

show abstract

Section: Combining Learning-based and Rule-based Decision Methodsmentioning

confidence: 99%

Driving Space for Autonomous Vehicles

et al. 2019

View full text Add to dashboard Cite

show abstract

“…In the autonomous driving domain, Hubschneider et al combine an end-to-end trained DNN proposing trajectories to a particle swarm optimizer (PSO) [14]. The DNN is trained via imitation learning using visual input from a front-facing camera and steering angle labels generated from an expert driver.…”

Section: B Learned Heuristicsmentioning

confidence: 99%

Decentralized Cooperative Planning for Automated Vehicles with Continuous Monte Carlo Tree Search

Kurzer

Engelhorn

Zöllner

2018

2018 21st International Conference on Intelligent Transportation Systems (ITSC)

Self Cite

View full text Add to dashboard Cite

Efficient driving in urban traffic scenarios requires foresight. The observation of other traffic participants, and the inference of their possible next actions depending on the own action is considered cooperative prediction and planning. Humans are well equipped with the capability to predict the actions of multiple interacting traffic participants and plan accordingly, without the need to directly communicate with others. Prior work has shown that it is possible to achieve effective cooperative planning without the need for explicit communication. However, the search space for cooperative plans is so large that the vast amount of the computational budget is spent on exploring the search space in unpromising regions that are far away from the solution. To accelerate the planning process, we combined learned heuristics with a cooperative planning method in order to guide the search towards regions with promising actions, yielding better results at lower computational costs.

show abstract

“…In addition, a learning-based approach was also introduced to design the motion planning algorithm. End to end approach was used to determine the desired motion directly from the measurement of sensors, such as a camera or laser scanner [41][42][43][44][45]. Hubschneider et al proposed the steering angle planning algorithm based on the Convolutional Neural Networks (CNN) [43].…”

Section: Introductionmentioning

confidence: 99%

“…End to end approach was used to determine the desired motion directly from the measurement of sensors, such as a camera or laser scanner [41][42][43][44][45]. Hubschneider et al proposed the steering angle planning algorithm based on the Convolutional Neural Networks (CNN) [43]. Similarly, Bai et al used Convolutional LSTM and 3D-CNN to determine the steering input [44].…”

Section: Introductionmentioning

confidence: 99%

Self-Adaptive Motion Prediction-Based Proactive Motion Planning for Autonomous Driving in Urban Environments

Jeong

2021

IEEE Access

View full text Add to dashboard Cite

This paper presents self-adaptive motion prediction-based proactive motion planning for autonomous driving in urban environments. In order to achieve fully autonomous driving in urban environments, the proposed algorithm predicts future behavior of moving vehicles and considers the potential risk of objects appearing suddenly from occluded regions. A self-adaptive motion predictor was used to predict the probabilistic future states of vehicles and estimate the uncertainty of prediction simultaneously. Then, a free space boundary and drivable corridor were defined to determine the desired longitudinal acceleration and path. Based on these two boundary definitions, the proposed proactive motion planner decided the desired longitudinal motion by considering the risk potential of objects, such as the appearance of vulnerable road users from the free space boundary. The desired path is determined within the drivable corridor by the proposed motion planner with an integrated motion optimizer. The performance of the proposed algorithm has been validated via vehicle tests. The test results demonstrated that the proposed algorithm proactively determined lateral and longitudinal motion to minimize risks caused by detected and possible objects simultaneously on complex narrow roads to ensure safety.

show abstract

Integrating end-to-end learned steering into probabilistic autonomous driving

Cited by 30 publications

References 12 publications

Driving Space for Autonomous Vehicles

Driving Space for Autonomous Vehicles

Decentralized Cooperative Planning for Automated Vehicles with Continuous Monte Carlo Tree Search

Self-Adaptive Motion Prediction-Based Proactive Motion Planning for Autonomous Driving in Urban Environments

Contact Info

Product

Resources

About