TTC-SLSTM: Human trajectory prediction using time-to-collision interaction energy - Intelligence Collective et Interaction Access content directly
Conference Papers Year : 2023

TTC-SLSTM: Human trajectory prediction using time-to-collision interaction energy

Abstract

Recent data-driven approaches using neural networks have shown promising results for pedestrian trajectory prediction. These algorithms outperform the knowledge-based and physics-based models in terms of distance error. However, it has been observed that the neural networks produce too many collisions and pedestrian overlaps, leading to unrealistic predictions. To address this problem, we propose in this contribution a hybrid extension of the Social-LSTM data-driven approach by introducing a collision loss in the training. The collision loss is provided by an interaction energy based on the timeto-collision with the neighbors. The predictions are evaluated and compared to the Social-LSTM model using both distanceerror metrics and collision metrics. The results show that collision and pedestrian overlap in the predicted trajectories decreases exponentially as the collision loss weight in the training increases, while the displacement error remains approximately constant. These preliminary results make the proposed hybrid algorithm a promising approach for realistic pedestrian prediction, especially in high-density situations.
Fichier principal
Vignette du fichier
KSE 2023 - vFinale.pdf (2.59 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04251961 , version 1 (20-10-2023)

Identifiers

  • HAL Id : hal-04251961 , version 1

Cite

Tu Dang-Huu, Raphael Korbmacher, Antoine Tordeux, Benoit Gaudou, Nicolas Verstaevel. TTC-SLSTM: Human trajectory prediction using time-to-collision interaction energy. 15th IEEE International Conference on Knowledge and Systems Engineering (KSE 2023), Academy of Cryptography Techniques (ACT); VNU University of Engineering and Technology (VNU-UET); IEEE, Oct 2023, Hanoi, Vietnam. à paraître. ⟨hal-04251961⟩
109 View
39 Download

Share

Gmail Facebook X LinkedIn More