Skip to content
DraconiticResearch
Search Sources API
203,203 studies31 sourcesUpdated 31 min ago
Peer-reviewedCC BY 4.0PLOS One

A non-local multi-head spatiotemporal attention LSTM for vehicle trajectory prediction

  • Sidra Rashid1
  • Muazzam A. Khan2
  • Usman Akram3
  • Raheel Mumtaz4
  • It Ee Lee5
  • Toqeer Ali Syed6
6 affiliations
  1. 1Department of Software Engineering, Bahria University, Islamabad, Pakistan
  2. 2ICESCO Chair Big Data Analytics and Edge Computing, Quaid-i-Azam University, Islamabad, Pakistan
  3. 3Department of Computer Sciences, Quaid-i-Azam University, Islamabad, Pakistan
  4. 4Department of Computer and Software Engineering, NUST CE & ME, Islamabad, Pakistan
  5. 5Faculty of Artificial Intelligence and Engineering, Multimedia University, Cyberjaya, Malaysia
  6. 6Faculty of Computer and Information System, Islamic University of Madinah, Madinah, Saudi Arabia
Open at sourcePDFDOICite

Abstract

Vehicle trajectory prediction (VTP) is the primary part of the perception-planning-control pipeline in autonomous driving. The inter-vehicle interactions strongly shape the trajectory of vehicles in complex traffic scenarios. Neural networks that use recurrent or convolutional architectures often exhibit performance degradation in longer prediction horizons. Many existing approaches confine interaction modeling to a predefined spatial neighborhood and a narrow temporal window, while implicitly assuming that all neighboring vehicles have equal influence on the future trajectory of target vehicle. This results in error accumulation and the inability to capture long-range spatiotemporal dependencies. To overcome these limitations, a non-local multi-head spatiotemporal attention based long short-term memory model (NL-MHA-LSTM) is introduced which employs an attention mechanism to assign context weights to relevant neighbor vehicles. It extends beyond pairwise effects to model long-range dependencies. The model emphasizes the most influential vehicles and formulate position-aware interaction representations. A comprehensive set of experiments is conducted on the pubicly available HighD dataset. The results demonstrate that the proposed model outperforms all state-of-the-art methods, achieving a 53.4% reduction in RMSE at the 5s prediction horizon relative to the next-best model. In addition, a detailed ablation study is conducted to systematically evaluate the impact of different attention mechanisms and varying numbers of attention heads on prediction accuracy across multiple horizons.

Related

  1. Hierarchical cross-modal spatiotemporal attention network for crop yield prediction in complex Himalayan agroecosystems
    Shabir Ahmad Mir, Umarfarooque Momin, Sudeep Kumar E. +3
    Frontiers in Plant SciencePeer-reviewedCC BY 4.0Full textPlant biologyCrossrefEurope PMCPMC
  2. Trajectory tracking control of unmanned aerial vehicle with disturbance observer and robust adaptive neural dynamic surface
    Xian Pan, Dongxue Wang, Youwu Liu +1
    PLOS OnePeer-reviewedCC BY 4.0Full textNeuroscienceAtmospheric science & meteorologyPLOS
  3. Cross-attention encoding models reveal dynamic spatiotemporal routing across human higher visual cortex
    Iishaan Inabathini, Margaret M. Henderson
    arXivPreprintCC BY-NC-ND 4.0Full textNeurosciencearXiv
  4. MSCF-DM: center-site-aware attention and gated multi-scale fusion network for cross-species DNA methylation site prediction
    Xin Wei, Siqin Hu
    BMC GenomicsPeer-reviewedCC BY 4.0Genetics & genomicsCrossref
  5. Longitudinal treatment selection under irreversible discontinuation: History-aware prediction and trajectory-aware learning in peanut-allergy prevention
    R. Taub, T. Wasserman, Y. Savir
    bioRxivPreprintCC BY-NCFull textBioinformatics & computational biologybioRxiv

Cite

BibTeXRIS
BibTeX
@article{rashid2026nonlocal,
  title = {{A non-local multi-head spatiotemporal attention LSTM for vehicle trajectory prediction}},
  author = {Sidra Rashid and Muazzam A. Khan and Usman Akram and Raheel Mumtaz and It Ee Lee and Toqeer Ali Syed},
  journal = {PLOS One},
  year = {2026},
  month = sep,
  doi = {10.1371/journal.pone.0357735},
  url = {https://doi.org/10.1371/journal.pone.0357735},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - A non-local multi-head spatiotemporal attention LSTM for vehicle trajectory prediction
AU  - Sidra Rashid
AU  - Muazzam A. Khan
AU  - Usman Akram
AU  - Raheel Mumtaz
AU  - It Ee Lee
AU  - Toqeer Ali Syed
PY  - 2026
DA  - 2026/09/21
JO  - PLOS One
DO  - 10.1371/journal.pone.0357735
UR  - https://doi.org/10.1371/journal.pone.0357735
AB  - Vehicle trajectory prediction (VTP) is the primary part of the perception-planning-control pipeline in autonomous driving. The inter-vehicle interactions strongly shape the trajectory of vehicles in complex traffic scenarios. Neural networks that use recurrent or convolutional architectures often exhibit performance degradation in longer prediction horizons. Many existing approaches confine interaction modeling to a predefined spatial neighborhood and a narrow temporal window, while implicitly assuming that all neighboring vehicles have equal influence on the future trajectory of target vehicle. This results in error accumulation and the inability to capture long-range spatiotemporal dependencies. To overcome these limitations, a non-local multi-head spatiotemporal attention based long short-term memory model (NL-MHA-LSTM) is introduced which employs an attention mechanism to assign context weights to relevant neighbor vehicles. It extends beyond pairwise effects to model long-range dependencies. The model emphasizes the most influential vehicles and formulate position-aware interaction representations. A comprehensive set of experiments is conducted on the pubicly available HighD dataset. The results demonstrate that the proposed model outperforms all state-of-the-art methods, achieving a 53.4% reduction in RMSE at the 5s prediction horizon relative to the next-best model. In addition, a detailed ablation study is conducted to systematically evaluate the impact of different attention mechanisms and varying numbers of attention heads on prediction accuracy across multiple horizons.
SN  - 1932-6203
ER  -