G-PECNet: Towards a Generalizable Pedestrian Trajectory Prediction System

TL;DR

G-PECNet is a deep generative model that improves pedestrian trajectory prediction accuracy for autonomous drone navigation by combining architectural enhancements, synthetic data augmentation, and a new non-linearity metric.

Contribution

It introduces G-PECNet, a novel model with architectural improvements and synthetic data augmentation techniques for better out-of-domain trajectory prediction.

Findings

9.5% improvement in Final Displacement Error on PECNet benchmark

Effective use of periodic activation functions and synthetic data augmentation

Proposed geometry-inspired metric for trajectory analysis

Abstract

Navigating dynamic physical environments without obstructing or damaging human assets is of quintessential importance for social robots. In this work, we solve autonomous drone navigation's sub-problem of predicting out-of-domain human and agent trajectories using a deep generative model. Our method: General-PECNet or G-PECNet observes an improvement of 9.5\% on the Final Displacement Error (FDE) on 2020's benchmark: PECNet through a combination of architectural improvements inspired by periodic activation functions and synthetic trajectory (data) augmentations using Hidden Markov Models (HMMs) and Reinforcement Learning (RL). Additionally, we propose a simple geometry-inspired metric for trajectory non-linearity and outlier detection, helpful for the task. Code available at https://github.com/Aryan-Garg/PECNet-Pedestrian-Trajectory-Prediction.git