Pedestrian trajectory prediction by supervised machine-learning
Classical operational pedestrian models rely on physical, social, or psychological factors. They are defined by basic rules or generic functions depending on the environment locally. These models are specified by few parameters adjusting the dynamics and that can be interpreted. Nowadays, machine learning and data-based algorithms such as artificial neural networks (ANN) are used for pedestrian trajectory prediction. Potential applications are pedestrian movement in complex situations, including evacuations, autonomous vehicles, or motion planning of robots in crowded environments.
G Saporta. Models for Understanding versus Models for Prediction. COMPSTAT 2008, pp. 315-322, 2008. Presentation
A Alahi, K Goel, V Ramanathan, A Robicquet, L Fei-Fei, S Savarese. Social LSTM: Human trajectory prediction in crowded spaces. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 961-971, 2016.
A Gupta, J Johnson, L Fei-Fei, S Savarese, A Alahi. Social GAN: Socially acceptable trajectories with generative adversarial networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2255-2264, 2018.
A Tordeux, M Chraibi, A Seyfried, A Schadschneider. Prediction of pedestrian speed with artificial neural networks. In International Conference on Traffic and Granular Flow, pp. 327-335, 2017 Poster
A Tordeux, M Chraibi, A Seyfried, A Schadschneider. Artificial neural networks predicting pedestrian dynamics in complex buildings. In SMSA2019: Stochastic Models, Statistics and their Application 2019, 363-372. Springer Proceedings in Mathematics & Statistics, vol 294. Springer, Cham, 2019. Presentation
A Tordeux, M Chraibi, A Seyfried, A Schadschneider. Prediction of pedestrian dynamics in complex architecture with artificial neural networks. Journal of Intelligent Transportation Systems, 24(6):556-568, 2019
R Korbmacher, A Tordeux. Review of Pedestrian Trajectory Prediction Methods: Comparing Deep Learning and Knowledge-based Approaches. arXiv preprint arXiv:2111.06740, 2021