FSTAN: A Two-Stage Traffic Conflict Prediction Framework Based on Vehicle Trajectory

被引:0
|
作者
Liu, Zhenyu [1 ]
Xiang, Qiaojun [1 ]
机构
[1] Southeast Univ, Sch Transportat, Jiangsu Key Lab Urban ITS, Nanjing 211189, Jiangsu, Peoples R China
基金
中国国家自然科学基金;
关键词
Trajectory; Predictive models; Attention mechanisms; Transformers; Force; Uncertainty; Spatiotemporal phenomena; Autonomous vehicles; Accuracy; Visualization; Field force attention; spatiotemporal transformer block (STTB); traffic conflict recognition; trajectory prediction; vehicle geometric bounding box; COLLISION-AVOIDANCE;
D O I
10.1109/TIM.2025.3533660
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Microscopic traffic conflict prediction is critical to improving road safety. For autonomous driving in highway scenarios, we develop a two-stage conflict prediction framework. In the first stage, a field force-based spatiotemporal attention network field-based spatial-temporal attention network (FSTAN) for trajectory prediction is proposed, which models vehicle interaction perception as a measurement process. The FSTAN network applies spatiotemporal transformer blocks (STTBs) with multiple attention mechanism to capture complex vehicle interactions. A potential field force attention mechanism is designed to capture spatial interactions, while self-attention mechanism is applied to model temporal interactions. The field force attention mechanism imparts physical significance to spatial interactions through field risk forces, enhancing the model's interpretability. In the second stage, a bounding box-based conflict recognition method is introduced, which takes into account the influence of vehicle geometry on conflict recognition. Experimental results highlight the superior long-term prediction capability of FSTAN compared to several state-of-the-art (SOTA) models. Visualization results demonstrate that the bounding box-based measurement method outperforms the center point-based method in both side-on and rear-end conflict detection. Center point-based measurement methods tend to underestimate collision risks and miss potential conflicts by neglecting vehicle geometry. The proposed two-stage framework lays a foundation for further development of real-world vehicle warning systems.
引用
收藏
页数:13
相关论文
共 50 条
  • [31] Two-Stage Segmentation Framework Based on Distance Transformation
    Huang, Xiaoyang
    Lin, Zhi
    Jiao, Yudi
    Chan, Moon-Tong
    Huang, Shaohui
    Wang, Liansheng
    SENSORS, 2022, 22 (01)
  • [32] Vision-Based Traffic Conflict Detection Using Trajectory Learning and Prediction
    Sun, Zongyuan
    Chen, Yuren
    Wang, Pin
    Fang, Shouen
    Tang, Boming
    IEEE ACCESS, 2021, 9 : 34558 - 34569
  • [33] Boosting Advanced Nasopharyngeal Carcinoma Stage Prediction Using a Two-Stage Classification Framework Based on Deep Learning
    Jin Huang
    Ruhan He
    Jia Chen
    Song Li
    Yuqin Deng
    Xinglong Wu
    International Journal of Computational Intelligence Systems, 14
  • [34] A Two-Stage Prediction Framework for Oil and Gas Well Production Based on Classification and Regression Models
    Hou, Dongdong
    Niu, Wente
    Han, Guoqing
    Sun, Yuping
    Zhang, Mingshan
    Liang, Xingyuan
    ENERGY & FUELS, 2024, 38 (22) : 22219 - 22229
  • [35] Two-stage hemoglobin prediction based on prior causality
    Chen, Yuwen
    Zhong, Kunhua
    Zhu, Yiziting
    Sun, Qilong
    FRONTIERS IN PUBLIC HEALTH, 2022, 10
  • [36] Social traits and credit card default: a two-stage prediction framework
    Gaganis, Chrysovalantis
    Papadimitri, Panagiota
    Pasiouras, Fotios
    Tasiou, Menelaos
    ANNALS OF OPERATIONS RESEARCH, 2023, 325 (02) : 1231 - 1253
  • [37] Social traits and credit card default: a two-stage prediction framework
    Chrysovalantis Gaganis
    Panagiota Papadimitri
    Fotios Pasiouras
    Menelaos Tasiou
    Annals of Operations Research, 2023, 325 : 1231 - 1253
  • [38] Evaluation of a two-stage framework for prediction using big genomic data
    Jiang, Xia
    Neapolitan, Richard E.
    BRIEFINGS IN BIOINFORMATICS, 2015, 16 (06) : 912 - 921
  • [39] A Two-Stage Active Domain Adaptation Framework for Vehicle Re-Identification
    Shang, Linzhi
    Zhao, Dawei
    Nie, Yiming
    Zhao, Kunlong
    Xiao, Liang
    Dai, Bin
    PATTERN RECOGNITION AND COMPUTER VISION, PRCV 2024, PT 1, 2025, 15031 : 380 - 394
  • [40] Two-stage vehicle availability model
    COM-SIS Corp, Silver Spring, United States
    Transp Res Rec, 1556 (16-21):