Pedestrian Trajectory Prediction with Convolutional Neural Networks

TL;DR

This paper introduces a novel 2D convolutional neural network model for pedestrian trajectory prediction, outperforming recurrent models and achieving state-of-the-art results on benchmark datasets, while also exploring social interaction modeling.

Contribution

The paper presents a new convolutional approach for pedestrian trajectory prediction, along with data augmentation techniques and an analysis of social information modeling.

Findings

The convolutional model outperforms recurrent models on ETH and TrajNet datasets.

Data augmentation techniques improve model robustness.

Occupancy-based social interaction modeling was found ineffective.

Abstract

Predicting the future trajectories of pedestrians is a challenging problem that has a range of application, from crowd surveillance to autonomous driving. In literature, methods to approach pedestrian trajectory prediction have evolved, transitioning from physics-based models to data-driven models based on recurrent neural networks. In this work, we propose a new approach to pedestrian trajectory prediction, with the introduction of a novel 2D convolutional model. This new model outperforms recurrent models, and it achieves state-of-the-art results on the ETH and TrajNet datasets. We also present an effective system to represent pedestrian positions and powerful data augmentation techniques, such as the addition of Gaussian noise and the use of random rotations, which can be applied to any model. As an additional exploratory analysis, we present experimental results on the inclusion of occupancy methods to model social information, which empirically show that these methods are ineffective in capturing social interaction.