EUVIMEDEUVIMED entdecken
Crossref

GTransPDM: A Graph-Embedded Transformer With Positional Decoupling for Pedestrian Crossing Intention Prediction

Chen Xie, Ciyun Lin, Xiaoyu Zheng, Bowen Gong, Antonio M. López

IEEE Signal Processing Letters · 2025 · 32 · 2109-2113

DOI: 10.1109/lsp.2025.3567249 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2025. Verfasst von Chen Xie und 4 weiteren Autor:innen. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 34 Literaturangaben. Gefördert von: Scientific and Technological Developing Project of Jilin Province, Generalitat de Catalunya CERCA Program and its ACCIO agency to the Computer Vision Center's general activities.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Understanding and predicting pedestrian crossing behavioral intention is crucial for the driving safety of autonomous vehicles. Nonetheless, challenges emerge when using promising images or environmental context masks to extract various factors for time-series network modeling, causing pre-processing errors or a loss of efficiency. Typically, pedestrian positions captured by onboard cameras are often distorted and do not accurately reflect their actual movements. To address these issues, GTransPDM – a Graph-embedded Transformer with a Position Decoupling Module – was developed for pedestrian crossing intention prediction by leveraging multi-modal features. First, a positional decoupling module was proposed to decompose pedestrian lateral motion and encode depth cues in the image view. Then, a graph-embedded Transformer was designed to capture the spatio-temporal dynamics of human pose skeletons, integrating essential factors such as position, skeleton, and ego-vehicle motion. Experimental results indicate that the proposed method achieves 92% accuracy on the PIE dataset and 87% accuracy on the JAAD dataset, with a processing speed of 0.05 ms. It outperforms the state-of-the-art in comparison.

Förderung

  • Scientific and Technological Developing Project of Jilin Province
  • Generalitat de Catalunya CERCA Program and its ACCIO agency to the Computer Vision Center's general activities

Zitieren

APA

Xie, C., Lin, C., Zheng, X., Gong, B., & López, A. M. (2025). GTransPDM: A Graph-Embedded Transformer With Positional Decoupling for Pedestrian Crossing Intention Prediction. IEEE Signal Processing Letters, 32, 2109-2113. https://doi.org/10.1109/lsp.2025.3567249

Vancouver

Xie C, Lin C, Zheng X, Gong B, López AM. GTransPDM: A Graph-Embedded Transformer With Positional Decoupling for Pedestrian Crossing Intention Prediction. IEEE Signal Processing Letters. 2025;32:2109-2113. doi:10.1109/lsp.2025.3567249

BibTeX
@article{Xie2025,
  author = {Xie, Chen and Lin, Ciyun and Zheng, Xiaoyu and Gong, Bowen and López, Antonio M.},
  title = {GTransPDM: A Graph-Embedded Transformer With Positional Decoupling for Pedestrian Crossing Intention Prediction},
  journal = {IEEE Signal Processing Letters},
  year = {2025},
  volume = {32},
  pages = {2109--2113},
  doi = {10.1109/lsp.2025.3567249}
}
Kostenlos starten →