Convolutional Social Pooling for Vehicle Trajectory Prediction

TL;DR

This paper introduces a convolutional social pooling method within an LSTM encoder-decoder framework to improve vehicle trajectory prediction by capturing interdependencies among neighboring vehicles, demonstrating superior accuracy on real traffic datasets.

Contribution

The paper presents a novel convolutional social pooling layer for vehicle trajectory prediction, enhancing interdependency learning and enabling multi-modal future trajectory distributions.

Findings

Improved RMS prediction error over state-of-the-art methods.

Lower negative log-likelihoods indicating better probabilistic predictions.

Qualitative analysis shows accurate modeling of various traffic scenarios.

Abstract

Forecasting the motion of surrounding vehicles is a critical ability for an autonomous vehicle deployed in complex traffic. Motion of all vehicles in a scene is governed by the traffic context, i.e., the motion and relative spatial configuration of neighboring vehicles. In this paper we propose an LSTM encoder-decoder model that uses convolutional social pooling as an improvement to social pooling layers for robustly learning interdependencies in vehicle motion. Additionally, our model outputs a multi-modal predictive distribution over future trajectories based on maneuver classes. We evaluate our model using the publicly available NGSIM US-101 and I-80 datasets. Our results show improvement over the state of the art in terms of RMS values of prediction error and negative log-likelihoods of true future trajectories under the model's predictive distribution. We also present a qualitative analysis of the model's predicted distributions for various traffic scenarios.