Physics-Informed Trajectory Prediction for Autonomous Driving under Missing Observation

Authors: Haicheng Liao, Chengyue Wang, Zhenning Li, Yongkang Li, Bonan Wang, Guofa Li, Chengzhong Xu

IJCAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive evaluations demonstrate that our approach markedly outperforms existing methods, achieving high accuracy even in scenarios with up to 75% missing observations.
Researcher Affiliation Academia 1University of Macau 2UESTC 3Chongqing University
Pseudocode Yes Algorithm 1 Risk Aware Algorithm
Open Source Code No The paper does not include an unambiguous statement or direct link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes We significantly enrich trajectory prediction research by introducing Mo CAD-missing, a dataset curated from a fully (Level 5) AV in real-world traffic scenarios, and by enhancing the established NGSIM and High D datasets with scenarios featuring missing observations.
Dataset Splits Yes These datasets provide longitudinal and lateral coordinates of traffic agents, which are systematically divided into training, validation, and test sets.
Hardware Specification Yes The model is trained on a single Nvidia A40 48GB GPU to convergence, using a learning rate of 0.0005 and a batch size of 64.
Software Dependencies Yes We use the Haar wavelet and its corresponding filters from Py Wavelet [Lee et al., 2019]
Experiment Setup Yes The model is trained on a single Nvidia A40 48GB GPU to convergence, using a learning rate of 0.0005 and a batch size of 64. We employ Mean Square Error (MSE) and Negative Log Likelihood (NLL) as loss functions.