bsuirДоклады БГУИРDoklady BGUIR1729-76482708-0382БГУИР10.35596/1729-7648-2025-23-4-109-117bsuir-4189Research ArticleСтатьиДинамическое графовое моделирование для многоагентного предсказания траектории движенияDynamic Relational Graph Modeling for Multi-Agent Motion Trajectory PredictionТанИ.TangYi

Минск 

Yi Tang, Postgraduate at the Department of Information Technologies in Automated Systems 

220013, Minsk, Brovki St., 6 

Tel.: +375 25 764-41-91 

tangyijcb@163.comПерцевД. Ю.PertsauD. Yu.

Минск 

Cand. Sci. (Tech.), Associate Professor, Associate Professor at the Department of Electronic Computer Science 

Minsk 

tangyijcb@163.comБелорусский государственный университет информатики и радиоэлектроникиBelarusian State University of Informatics and Radioelectronics202503092025234109117Copyright © Тан И., Перцев Д.Ю., 20252025Тан И., Перцев Д.Ю.Tang Y., Pertsau D.Y.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://doklady.bsuir.by/jour/article/view/4189

Точное прогнозирование траектории движения нескольких агентов является важнейшей задачей в таких областях, как автономное вождение, взаимодействие человека с компьютером и анализ поведения. Однако динамичность и интерактивность поведения агентов создают значительные проблемы, поскольку требуют формирования сложных пространственно-временных зависимостей и динамичес­ ки развивающегося взаимодействия между агентами. Предлагается новый подход для моделирования динамических реляционных графов, основным компонентом которых является блок акцента внимания с учетом относительного положения агентов на основе графов. Рассматривая объекты в сцене (например, транспортные средства и элементы дороги) как узлы графа, а их взаимодействие как ребра, предложенный подход эффективно отражает как локальные, так и глобальные зависимости на сцене и делает прогноз о будущей траектории. Представленный подход оценивается с помощью набора данных для прогнозирования траектории Argoverse1. Экспериментальные результаты показывают, что такая модель превосходит существующие методы.

Accurate trajectory prediction of multiple agents is a critical task in the fields of autonomous driving, human-computer interaction, and behavior analysis. However, the dynamic and interactive nature of agent behavior poses significant challenges, since it requires the formation of complex spatio-temporal dependencies and dynamically evolving interactions between agents. A novel approach is proposed for modeling dynamic relational graphs, the core component of which is the attention focus block, taking into account the relative positions of graph-based agents. By considering objects in a scene (e.g., vehicles and road elements) as graph nodes and their interactions as edges, the proposed approach effectively captures both local and global dependencies in a scene and makes a prediction about the future trajectory. The presented approach is evaluated using the Argoverse1 trajectory prediction dataset. Experimental results show that this model outperforms existing methods.

многоагентное прогнозирование траекторииграфовая нейронная сетьмеханизм вниманияmulti-agent trajectory predictiongraph neural networkattention blockReferencesYurtsever E., Lambert J., Carballo A., Takeda K. (2020) A Survey of Autonomous Driving: Common Practi­ ces and Emerging Technologies. IEEE Access. 8, 58443–58469.Yurtsever E., Lambert J., Carballo A., Takeda K. (2020) A Survey of Autonomous Driving: Common Practi­ ces and Emerging Technologies. IEEE Access. 8, 58443–58469.Wang S., Bao Z., Culpepper J. S., Cong G. (2021) A Survey on Trajectory Data Management, Analytics, and Learning. ACM Computing Surveys (CSUR). 54 (2), 1–36.Wang S., Bao Z., Culpepper J. S., Cong G. (2021) A Survey on Trajectory Data Management, Analytics, and Learning. ACM Computing Surveys (CSUR). 54 (2), 1–36.Singh A. (2023) Trajectory-Prediction with Vision: A Survey. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 3318–3323.Singh A. (2023) Trajectory-Prediction with Vision: A Survey. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 3318–3323.Casas S., Luo W., Urtasun R. (2018) IntentNet: Learning to Predict Intention from Raw Sensor Data. In Proceedings of the Conference on Robot Learning. Zürich, Switzerland. 947–956.Casas S., Luo W., Urtasun R. (2018) IntentNet: Learning to Predict Intention from Raw Sensor Data. In Proceedings of the Conference on Robot Learning. Zürich, Switzerland. 947–956.Hong J., Sapp B., Philbin J. (2019) Rules of the Road: Predicting Driving Behavior with a Convolutional Mo­ del of Semantic Interactions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, CA, USA. 8454–8462.Hong J., Sapp B., Philbin J. (2019) Rules of the Road: Predicting Driving Behavior with a Convolutional Mo­ del of Semantic Interactions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, CA, USA. 8454–8462.Gao J., Sun C., Zhao H., Shen Y., Anguelov D., Li C., et al. (2020) VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 11525–11533.Gao J., Sun C., Zhao H., Shen Y., Anguelov D., Li C., et al. (2020) VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 11525–11533.Liang M., Yang B., Hu R., Chen Y., Liao R., Feng S., et al. (2020) Learning Lane Graph Representations for Motion Forecasting. In Computer Vision – ECCV 2020: 16th European Conference, Glasgow, UK, Aug. 23–28, 2020, Proceedings, Part II. Springer International Publishing. 541–556.Liang M., Yang B., Hu R., Chen Y., Liao R., Feng S., et al. (2020) Learning Lane Graph Representations for Motion Forecasting. In Computer Vision – ECCV 2020: 16th European Conference, Glasgow, UK, Aug. 23–28, 2020, Proceedings, Part II. Springer International Publishing. 541–556.Zhou Z., Ye L., Wang J., Wu K., Lu K. (2022) HiVT: Hierarchical Vector Transformer for Multi-Agent Motion Prediction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8823–8833.Zhou Z., Ye L., Wang J., Wu K., Lu K. (2022) HiVT: Hierarchical Vector Transformer for Multi-Agent Motion Prediction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8823–8833.Zhang L., Li P., Liu S., Shen S. (2024) SIMPL: A Simple and Efficient Multi-Agent Motion Prediction Baseline for Autonomous Driving. IEEE Robotics and Automation Letters.Zhang L., Li P., Liu S., Shen S. (2024) SIMPL: A Simple and Efficient Multi-Agent Motion Prediction Baseline for Autonomous Driving. IEEE Robotics and Automation Letters.Qi C. R., Su H., Mo K., Guibas L. J. (2017) PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 652–660.Qi C. R., Su H., Mo K., Guibas L. J. (2017) PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 652–660.Chang M.-F., Lambert J., Sangkloy P., Singh J., Bak S., Hartnett A., et al. (2019) Argoverse: 3D Tracking and Forecasting with Rich Maps. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8748–8757.Chang M.-F., Lambert J., Sangkloy P., Singh J., Bak S., Hartnett A., et al. (2019) Argoverse: 3D Tracking and Forecasting with Rich Maps. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8748–8757.Gu J., Sun C., Zhao H. (2021) DenseTNT: End-to-End Trajectory Prediction from Dense Goal Sets. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 15303–15312.Gu J., Sun C., Zhao H. (2021) DenseTNT: End-to-End Trajectory Prediction from Dense Goal Sets. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 15303–15312.Gilles T., Sabatini S., Tsishkou D., Stanciulescu B., Moutarde F. (2022) THOMAS: Trajectory Heatmap Output with Learned Multi-Agent Sampling. In International Conference on Learning Representations.Gilles T., Sabatini S., Tsishkou D., Stanciulescu B., Moutarde F. (2022) THOMAS: Trajectory Heatmap Output with Learned Multi-Agent Sampling. In International Conference on Learning Representations.Wang M., Zhu X., Yu C., Li W., Ma Y., Jin R., et al. (2023) GANet: Goal Area Network for Motion Forecas­ ting. In 2023 IEEE International Conference on Robotics and Automation (ICRA). 1609–1615.Wang M., Zhu X., Yu C., Li W., Ma Y., Jin R., et al. (2023) GANet: Goal Area Network for Motion Forecas­ ting. In 2023 IEEE International Conference on Robotics and Automation (ICRA). 1609–1615.Ngiam J., Vasudevan V., Caine B., Zhang Z., Chiang H. T. L., Ling J., et al. (2023) Scene Transformer: A Unified Architecture for Predicting Future Trajectories of Multiple Agents. In International Conference on Learning Representations.Ngiam J., Vasudevan V., Caine B., Zhang Z., Chiang H. T. L., Ling J., et al. (2023) Scene Transformer: A Unified Architecture for Predicting Future Trajectories of Multiple Agents. In International Conference on Learning Representations.Feng C., Zhou H., Lin H., Zhang Z., Xu Z., Zhang C., et al. (2023) MacFormer: Map-Agent Coupled Transformer for Real-Time and Robust Trajectory Prediction. IEEE Robotics and Automation Letters.Feng C., Zhou H., Lin H., Zhang Z., Xu Z., Zhang C., et al. (2023) MacFormer: Map-Agent Coupled Transformer for Real-Time and Robust Trajectory Prediction. IEEE Robotics and Automation Letters.

The authors declare that there are no conflicts of interest present.