Pedestrian Path Prediction with Recursive Bayesian Filters: A Comparative Study

Abstract

Abstract. In the context of intelligent vehicles, we perform a comparative study on recursive Bayesian filters for pedestrian path prediction at short time horizons (< 2s). We consider Extended Kalman Filters (EKF) based on single dynamical models and Interacting Multiple Models (IMM) combining several such basic models (constant velocity/acceleration/turn). These are applied to four typical pedestrian motion types (crossing, stopping, bending in, starting). Position measurements are provided by an external state-of-the-art stereo vision-based pedestrian detector. We investigate the accuracy of position estimation and path prediction, and the benefit of the IMMs vs. the simpler single dynamical models. Special care is given to the proper sensor modeling and parameter optimization. The dataset and evaluation framework are made public to facilitate benchmarking.

Using road topology to improve cyclist path prediction

“…As in [3], [14], observations are filtered online using a recursive Bayesian filter with the selected motion model. At any frame, a predictive distribution for future positions is obtained by executing a filter's 'predict' step several times without any 'update'.…”

“…But predictive models of a pedestrian's path must represent spatial uncertainty too. A common approach leverages the motion models which are already an integral part of the tracker's filter [3]. More informed predictions are obtained by conditioning dynamics on additional cues, such as intent and awareness of the pedestrian [14], [15] or driver [16].…”

“…For temporal alignment, we follow existing literature [3], Fig. 3: Three tracks, before (left) and after (right) they have been transformed to the general coordinate system.…”

“…The following probabilistic motion models are considered: a) Linear Dynamical System (LDS): Previous research on pedestrian path prediction found no significant benefit of higher-order or constant turn motion models over a constant velocity model with white noise acceleration [3]. Hence we pick the constant velocity LDS as our baseline.…”

“…Most literature on VRU path prediction focuses on pedestrians (e.g. [3], [4], [5], [6]), where various cues have been proposed to improve trajectory prediction, such as pedestrian attention, spatial layout, etc. Predicting cyclist tracks from a moving vehicle is also challenging, as cyclists move fast, and can be observed for long durations where high-level behavior results in distinct paths, e.g.…”