[1]卢一凡,李煊鹏,薛启凡.面向未知域场景的车辆轨迹预测模型[J].智能系统学报,2024,19(5):1238-1247.[doi:10.11992/tis.202306046]
LU Yifan,LI Xuanpeng,XUE Qifan.Vehicle trajectory prediction model for unseen domain scenarios[J].CAAI Transactions on Intelligent Systems,2024,19(5):1238-1247.[doi:10.11992/tis.202306046]
点击复制
LU Yifan,LI Xuanpeng,XUE Qifan.Vehicle trajectory prediction model for unseen domain scenarios[J].CAAI Transactions on Intelligent Systems,2024,19(5):1238-1247.[doi:10.11992/tis.202306046]
面向未知域场景的车辆轨迹预测模型
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
19
期数:
2024年第5期
页码:
1238-1247
栏目:
学术论文—智能系统
出版日期:
2024-09-05
- Title:
- Vehicle trajectory prediction model for unseen domain scenarios
- Keywords:
- track prediction; domain generalization; invariant risk minimization; conditional variational auto encoder; endpoint generating; vector map; scene context; time-series networks
- 分类号:
- TP391
- 文献标志码:
- 2024-08-28
- 摘要:
- 自动驾驶技术随着科技革新迎来蓬勃发展，轨迹预测已成为智能汽车软件系统不可或缺的关键组成部分。为了解决传统车辆轨迹预测模型中存在的泛化能力不足的问题，提出一种基于泛化终点预测和地图场景的车辆轨迹预测方法。该方法采用基于不变风险最小化的条件变分自编码器生成轨迹终点，并结合时序网络编码的地图场景特征，提升了模型预测未知域数据的准确率。在交互式道路场景数据集INTERACTION上的实验结果证明该模型具有良好的泛化性能。本方法与效果最好的方法REx相比1 、2 、3 s处的mADE值(越小越好)分别下降0%、36.59%、50.68%，在未知测试域的预测轨迹准确度得到显著提升。
- Abstract:
- With the rapid development of autonomous driving technology, trajectory prediction has become an essential component of smart car software systems. To address the limitations in the generalization of traditional vehicle trajectory prediction models, this study proposes a vehicle trajectory prediction method based on generalized endpoint prediction and vector maps. This method employs a conditional variational autoencoder based on invariant risk minimization to generate trajectory endpoints. The accuracy of the model in predicting unseen domains is improved by integrating map scene features encoded by a time-series network. Experiments were conducted using the interaction dataset, which includes interactive driving scenarios. Experimental results showed that, compared with the best-performing state-of-the-art (SOTA) method REx, the mADE values (where lower is better) decreased by 0%, 36.59%, and 50.68% at 1, 2, and 3 s, respectively. The accuracy of the predicted trajectories in the unseen test domain was significantly improved.
- 参考文献/References:
-
[1] HUANG Yanjun, DU Jiatong, YANG Ziru, et al. A survey on trajectory-prediction methods for autonomous driving[J]. IEEE transactions on intelligent vehicles, 2022, 7(3): 652-674.
[2] 王翎, 孙涵. 基于域适应的多场景车辆检测[J]. 计算机技术与发展, 2019, 29(12): 158-161, 166.
WANG Ling, SUN Han. Vehicle detection in different scenes based on domain adaptation[J]. Computer technology and development, 2019, 29(12): 158-161, 166.
[3] ARJOVSKY M, BOTTOU L, GULRAJANI I, et al. Invariant risk minimization[EB/OL]. (2019-07-05) [2023-06-26]. https://arxiv.org/abs/1907.02893.
[4] ELMAN J L. Finding structure in time[J]. Cognitive science, 1990, 14(2): 179-211.
[5] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural computation, 1997, 9(8): 1735-1780.
[6] CHO K, VAN MERRIENBOER B, GULCEHRE C, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. Doha: ACL, 2014: 1724–1734.
[7] 柴进. 基于聚类和LSTM算法的车辆轨迹预测模型研究[D]. 北京: 北京交通大学, 2020.
CHAI Jin. On vehicle trajectory prediction model based on combined clustering and LSTM algorithm[D]. Beijing: Beijing Jiaotong University, 2020.
[8] 孔玮, 刘云, 李辉, 等. 基于深度学习的行人轨迹预测方法综述[J]. 控制与决策, 2021, 36(12): 2841-2850.
KONG Wei, LIU Yun, LI Hui, et al. Survey of pedestrian trajectory prediction methods based on deep learning[J]. Control and decision, 2021, 36(12): 2841-2850.
[9] PARK S H, KIM B, KANG C M, et al. Sequence-to-sequence prediction of vehicle trajectory via LSTM encoder-decoder architecture[C]//2018 IEEE Intelligent Vehicles Symposium. Changshu: IEEE, 2018: 1672-1678.
[10] 赵靖文, 李煊鹏, 张为公. 车辆多目标交互行为建模的轨迹预测方法[J]. 智能系统学报, 2023, 18(3): 480-488.
ZHAO Jingwen, LI Xuanpeng, ZHANG Weigong. Vehicle trajectory prediction method based on modeling of multi-agent interaction behavior[J]. CAAI transactions on intelligent systems, 2023, 18(3): 480-488.
[11] ALAHI A, GOEL K, RAMANATHAN V, et al. Social LSTM: human trajectory prediction in crowded spaces[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 961-971.
[12] DEO N, TRIVEDI M M. Convolutional social pooling for vehicle trajectory prediction[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Salt Lake City: IEEE, 2018: 1468-1476.
[13] 曾伟良, 陈漪皓, 姚若愚, 等. 时空图注意力网络在交叉口车辆轨迹预测的应用[J]. 计算机科学, 2021, 48(S1): 334-341.
ZENG Weiliang, CHEN Yihao, YAO Ruoyu, et al. Application of spatial-temporal graph attention networks in trajectory prediction for vehicles at intersections[J]. Computer science, 2021, 48(S1): 334-341.
[14] 赵靖文. 基于改进Transformer模型的车辆轨迹预测研究[D]. 南京: 东南大学, 2022.
ZHAO Jingwen. Research on vehicle trajectory prediction based on the improved transformer model. Nanjing: Southeast University, 2022.
[15] DIEHL F, BRUNNER T, LE M T, et al. Graph neural networks for modelling traffic participant interaction[C]//2019 IEEE Intelligent Vehicles Symposium. Paris: IEEE, 2019: 695-701.
[16] MOHAMED A, QIAN Kun, ELHOSEINY M, et al. Social-STGCNN: a social spatio-temporal graph convolutional neural network for human trajectory prediction[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 14412-14420.
[17] LI Xin, YING Xiaowen, CHUAH M C. GRIP: graph-based interaction-aware trajectory prediction[C]// 2019 IEEE Intelligent Transportation Systems Conference. Auckland: IEEE, 2019: 3960-3966.
[18] LI Xin, YING Xiaowen, CHUAH M C. GRIP++: enhanced graph-based interaction-aware trajectory prediction for autonomous driving[EB/OL]. (2019-07-17) [2023-06-26]. https://arxiv.org/abs/1907.07792.
[19] ZHAO Hang, GAO Jiyang, LAN Tian, et al. TNT: target-driven trajectory prediction[EB/OL]. (2020-08-19) [2023-06-26]. https://arxiv.org/abs/2008.08294.
[20] GU Junru, SUN Chen, ZHAO Hang. DenseTNT: end-to-end trajectory prediction from dense goal sets[C]//2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 15283-15292.
[21] LIN Yong, DONG Hanze, WANG Hao, et al. Bayesian invariant risk minimization[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 16000-16009.
[22] ZHOU Xiao, LIN Yong, ZHANG Weizhong, et al. Sparse invariant risk minimization[C]//Proceedings of the 39th International Conference on Machine Learning. Baltimore: PMLR, 2022: 27222-27244.
[23] ZIEGLER J, BENDER P, SCHREIBER M, et al. Making bertha drive—an autonomous journey on a historic route[J]. IEEE intelligent transportation systems magazine, 2014, 6(2): 8-20.
[24] HONG J, SAPP B, PHILBIN J. Rules of the road: predicting driving behavior with a convolutional model of semantic interactions[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 8454-8462.
[25] GAO Jiyang, SUN Chen, ZHAO Hang, et al. VectorNet: encoding HD maps and agent dynamics from vectorized representation[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 11522-11530.
[26] ZENG Wenyuan, LIANG Ming, LIAO Renjie, et al. LaneRCNN: distributed representations for graph-centric motion forecasting[C]//2021 IEEE/RSJ International Conference on Intelligent Robots and Systems. Prague: IEEE, 2021: 532-539.
[27] WANG Jindong, LAN Cuiling, LIU Chang, et al. Generalizing to unseen domains: a survey on domain generalization[J]. IEEE transactions on knowledge and data engineering, 2023, 35(8): 8052-8072.
[28] PENG Xi, QIAO Fengchun, ZHAO Long. Out-of-domain generalization from a single source: an uncertainty quantification approach[J]. IEEE transactions on pattern analysis and machine intelligence, 2024, 46(3): 1775-1787.
[29] SOMAVARAPU N , MA C Y , KIRA Z . Frustratingly simple domain generalization via image stylization[EB/OL]. (2020-06-19) [2023-06-26]. https://arxiv.org/abs/2006.11207.
[30] SEGU M, TONIONI A, TOMBARI F. Batch normalization embeddings for deep domain generalization[J]. Pattern recognition, 2023, 135: 109115.
[31] LI Da, YANG Yongxin, SONG Yizhe, et al. Learning to generalize: meta-learning for domain generalization[J]. Proceedings of the AAAI conference on artificial intelligence, 2018, 32(1): 3490-3497.
[32] GAN Chuang, YANG Tianbao, GONG Boqing. Learning attributes equals multi-source domain generalization[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 87-97.
[33] LI Haoliang, PAN S J, WANG Shiqi, et al. Domain generalization with adversarial feature learning[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 5400-5409.
[34] KRUEGER D, CABALLERO E, JACOBSEN J H, et al. Out-of-distribution generalization via risk extrapolation (REx)[EB/OL]. (2020-03-02) [2023-06-26]. https://arxiv.org/abs/2003.00688.
[35] 侯廉. 面向自动驾驶应用的周车轨迹群体交互式预测[D]. 北京: 清华大学, 2019.
HOU Lian. Distributed and interactive prediction of future trajectories of surrounding vehicles for autonomous driving. Beijing: Tsinghua University, 2019.
[36] 王少博. 动态场景下基于交互性预测的自动驾驶汽车轨迹规划方法研究[D]. 合肥: 中国科学技术大学, 2020.
WANG Shaobo. Research on interaction-aware predictive trajectory planning for autonomous vehicle in dynamic environments. Hefei: University of Science and Technology of China, 2020.
[37] ZHAN Wei, SUN Liting, WANG Di, et al. Interaction dataset: an international, adversarial and cooperative motion dataset in interactive driving scenarios with semantic maps[EB/OL]. (2019-10-30) [2023-06-26]. https://arxiv.org/abs/1910.03088.
[38] VAPNIK V N. Statistical learning theory[M]. New York: Wiley, 1998.
[39] GANIN Y , USTINOVA E , AJAKAN H, et al. Domain-adversarial training of neural networks[J]. Journal of machine learning research, 2016, 17(1): 2096-2030.
- 相似文献/References:
-
[1]赵靖文,李煊鹏,张为公.车辆多目标交互行为建模的轨迹预测方法[J].智能系统学报,2023,18(3):480.[doi:10.11992/tis.202201022]
ZHAO Jingwen,LI Xuanpeng,ZHANG Weigong.Vehicle trajectory prediction method based on modeling of multi-agent interaction behavior[J].CAAI Transactions on Intelligent Systems,2023,18():480.[doi:10.11992/tis.202201022]
备注/Memo
收稿日期:2023-6-26。
基金项目:国家重点研发计划项目(2021YFB1600501)；国家自然科学基金项目(61906038)；东南大学至善青年学者项目&中央高校基本科研业务费专项资金项目(2242021R41184).
作者简介:卢一凡，硕士研究生，主要研究方向为道路场景下的模型领域泛化。E-mail：18502517268@163.com;李煊鹏，副教授，博士，主要研究方向为道路场景下的环境感知、行为预测及因果推理。主持和参与国家级及省部级项目10余项，授权发明专利7项，发表学术论文30余篇。E-mail：li_xuanpeng@seu.edu.cn;薛启凡，博士研究生，主要研究方向为自动驾驶领域的轨迹预测和模型领域泛化。授权发明专利4项，发表学术论文 5篇。E-mail：xue_qifan@seu.edu.cn。
通讯作者:李煊鹏. E-mail：li_xuanpeng@seu.edu.cn
更新日期/Last Update:
2024-09-05