2022
DOI: 10.1016/j.aap.2022.106711
|View full text |Cite
|
Sign up to set email alerts
|

Can automated driving prevent crashes with distracted Pedestrians? An exploration of motion planning at unsignalized Mid-block crosswalks

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
3
0

Year Published

2022
2022
2025
2025

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 13 publications
(3 citation statements)
references
References 46 publications
0
3
0
Order By: Relevance
“…Zhu et al [ 78 ] attempted to create automated driving control models that can prevent safety hazards caused by distracted pedestrians with different crossing speeds at unsignalized mid-block crosswalks. They developed an original rule-based model that assumes all detected pedestrians are distracted and an original learning-based model using reinforcement learning methods with a comprehensive reward function that considers failure cases and acceptable minimum safety margin time for pedestrians.…”
Section: Marl For Cavsmentioning
confidence: 99%
“…Zhu et al [ 78 ] attempted to create automated driving control models that can prevent safety hazards caused by distracted pedestrians with different crossing speeds at unsignalized mid-block crosswalks. They developed an original rule-based model that assumes all detected pedestrians are distracted and an original learning-based model using reinforcement learning methods with a comprehensive reward function that considers failure cases and acceptable minimum safety margin time for pedestrians.…”
Section: Marl For Cavsmentioning
confidence: 99%
“…Tang et al [22] introduced a motion planning method for automated driving using a soft actor-critic method [23], in which different strategies were balanced via the weights of safety, comfort, and efficiency. Zhu et al [24] developed a motion planning method to consider the factor of pedestrian distraction based on a rule-based method and a learning-based method. Their experimental results demonstrated that the learning-based method may be better at handling the unsafe action problem at unsignalized mid-block crosswalks than the rule-based method; nonetheless, the learningbased method may generate unreasonable actions.…”
Section: Related Workmentioning
confidence: 99%
“…Research on detection and localization algorithms with robust target detection capabilities in complex natural environments is necessary. Deep-learning techniques have been widely used in visual detection [11][12][13][14][15]. In the field of target detection, mainstream target detection algorithms are usually categorized into single-stage target detection, e.g., the YOLO [16] (you only look once) series and SSD [17] (single shot multibox detector), and two-stage target detection, e.g., RCNN (Region CNN) [18], Fast RCNN [19], and Faster RCNN [20].…”
Section: Introductionmentioning
confidence: 99%