电子学报2025，Vol.53Issue(10)：3640-3658,19.DOI:10.12263/DZXB.20250729
分级包络对抗域适应和松-紧耦合行人轨迹预测模型
Hierarchical Envelope Adversarial Domain Adaptation with Loose-Tight Coupled Trajectory Prediction Model
摘要
Abstract
Pedestrian trajectory prediction holds significant applications across autonomous driving,intelligent sur-veillance,and smart cities.However,the complexity and unpredictability of pedestrian interactions make this task a persis-tent challenge.Current models face two common limitations:(1)Considering only a single type of social coupling.This in-troduces redundant interactions.More critically,it fails to account for the varying nature of trajectory coupling across differ-ent scenarios and between different pedestrians.Consequently,models cannot deeply explore or effectively utilize diverse scene features;(2)Inadequate handling of domain shift.Although very few methods address domain shift,they rely on sta-tistical criterion-based domain distribution alignment.Such approaches exhibit strong dependency on predefined statistical metrics.This leads to significant limitations in complex,dynamic environments.To address these issues,this paper propos-es a hierarchical envelope adversarial domain adaptation with loose-tight coupled model.Firstly,an envelope sample trans-formation mechanism was designed.It constructs envelope samples and extends them into graph structures;Secondly,an ad-versarial domain adaptation module was developed.This integrates both local and global domain adaptation strategies;mean-while,a loose-tight coupling envelope sample construction module was created.It dynamically adapts to diverse coupling re-lationships across scenarios.These innovations collectively enhance prediction accuracy and robustness.The experimental section employed two representative public datasets for validation and conducted comprehensive comparisons with six rele-vant baseline algorithms.Results demonstrate that our model achieves significantly higher accuracy compared to existing methods,with the average displacement error(ADE)and final displacement error(FDE)metrics reduced by 17.6%and 19.1%,respectively.The time overhead meets practical requirements,which verifies the effectiveness of our key innovations.关键词
行人轨迹预测/社会交互/对抗域适应/包络样本变换/松-紧耦合Key words
pedestrian trajectory prediction/social interaction/adversarial domain adaptation/envelope sample transformation/loose-tight coupled分类
信息技术与安全科学引用本文复制引用
李勇明,胡杰,张小恒,王品,李文正..分级包络对抗域适应和松-紧耦合行人轨迹预测模型[J].电子学报,2025,53(10):3640-3658,19.基金项目
国家自然科学基金(No.U21A20448,No.61771080,No.72001032) National Natural Science Foundation of China(No.U21A20448,No.61771080,No.72001032) （No.U21A20448,No.61771080,No.72001032）
