Improving the Generalizability of Trajectory Prediction Models with Frenet-Based Domain Normalization

被引:2
|
作者
Ye, Luyao [1 ,2 ]
Zhou, Zikang [1 ,2 ]
Wang, Jianping [1 ,2 ]
机构
[1] City Univ Hong Kong, Dept Comp Sci, Hong Kong, Peoples R China
[2] City Univ Hong Kong, Shenzhen Res Inst, Shenzhen, Peoples R China
来源
2023 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2023) | 2023年
关键词
D O I
10.1109/ICRA48891.2023.10160788
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Predicting the future trajectories of robots' nearby objects plays a pivotal role in applications such as autonomous driving. While learning-based trajectory prediction methods have achieved remarkable performance on public benchmarks, the generalization ability of these approaches remains questionable. The poor generalizability on unseen domains, a well-recognized defect of data-driven approaches, can potentially harm the real-world performance of trajectory prediction models. We are thus motivated to improve models' generalization ability instead of merely pursuing high accuracy on average. Due to the lack of benchmarks for quantifying the generalization ability of trajectory predictors, we first construct a new benchmark called argoverse-shift, where the data distributions of domains are significantly different. Using this benchmark for evaluation, we identify that the domain shift problem seriously hinders the generalization of trajectory predictors since state-of-the-art approaches suffer from severe performance degradation when facing those out-of-distribution scenes. To enhance the robustness of models against domain shift problem, we propose a plug-and-play strategy for domain normalization in trajectory prediction. Our strategy utilizes the Frenet coordinate frame for modeling and can effectively narrow the domain gap of different scenes caused by the variety of road geometry and topology. Experiments show that our strategy noticeably boosts the prediction performance of the state-of-the-art in domains that were previously unseen to the models, thereby improving the generalization ability of data-driven trajectory prediction methods.
引用
收藏
页码:11562 / 11568
页数:7
相关论文
共 50 条
  • [31] Utility of prediction model score: a proposed tool to standardize the performance and generalizability of clinical predictive models based on systematic review
    Ehresman, Jeff
    Lubelski, Daniel
    Pennington, Zach
    Hung, Bethany
    Ahmed, A. Karim
    Azad, Tej D.
    Lehner, Kurt
    Feghali, James
    Buser, Zorica
    Harrop, James
    Wilson, Jefferson
    Kurpad, Shekar
    Ghogawala, Zoher
    Sciubba, Daniel M.
    JOURNAL OF NEUROSURGERY-SPINE, 2021, 34 (05) : 779 - 787
  • [32] Improving Prediction in Genetic Models Based on Additive Effects from Dependent Variants
    Shedden, Kerby
    Hsu, Ming-Chi
    Zhu, Ji
    Sun, Yan V.
    GENETIC EPIDEMIOLOGY, 2010, 34 (08) : 980 - 980
  • [33] Ensemble component selection for improving ICA based microarray data prediction models
    Liu, Kun-Hong
    Li, Bo
    Zhang, Jun
    Du, Ji-Xiang
    PATTERN RECOGNITION, 2009, 42 (07) : 1274 - 1283
  • [34] Improving the performance of ICA based microarray data prediction models with genetic algorithm
    Liu, Kun-Hong
    Huang, De-Shuang
    Li, Bo
    2007 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-10, PROCEEDINGS, 2007, : 606 - +
  • [35] Domain-based predictive models for protein-protein interaction prediction
    Chen, Xue-Wen
    Liu, Mei
    Eurasip Journal on Applied Signal Processing, 2006, 2006
  • [36] Domain-Based Predictive Models for Protein-Protein Interaction Prediction
    Xue-Wen Chen
    Mei Liu
    EURASIP Journal on Advances in Signal Processing, 2006
  • [37] Domain-based predictive models for protein-protein interaction prediction
    Chen, Xue-Wen
    Liu, Mei
    EURASIP JOURNAL ON APPLIED SIGNAL PROCESSING, 2006, 2006 (1)
  • [38] Vehicle Trajectory-Prediction Method Based on Driver Behavior-Classification and Informer Models
    Su, Jianyu
    Li, Muyang
    Zhu, Langqian
    Zhang, Sijia
    Liu, Mingjian
    WORLD ELECTRIC VEHICLE JOURNAL, 2024, 15 (06):
  • [39] Hybrid deep learning models for ship trajectory prediction in complex scenarios based on AIS data
    Liu, Zhiheng
    Qi, Wenjuan
    Zhou, Suiping
    Zhang, Wenjie
    Jiang, Cheng
    Jie, Yongshi
    Li, Chengyang
    Guo, Yuru
    Guo, Jianhua
    APPLIED OCEAN RESEARCH, 2024, 153
  • [40] Spontaneous EEG-Based Normalization of Pain-evoked Neural Responses: Effect on Improving the Accuracy of Pain Prediction
    Bai, Yanru
    Hu, Yong
    Zhang, Zhiguo
    2016 IEEE INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE AND VIRTUAL ENVIRONMENTS FOR MEASUREMENT SYSTEMS AND APPLICATIONS (CIVEMSA), 2016, : 53 - 56