Title:	Trajectory Prediction Model Dataset Generalization
DNr:	Berzelius-2026-142
Project Type:	LiU Berzelius
Principal Investigator:	Erik Frisk <erik.frisk@liu.se>
Affiliation:	Linköpings universitet
Duration:	2026-05-30 – 2026-12-01
Classification:	20208
Homepage:	https://liu.se/medarbetare/erifr93
Keywords:

Abstract

This project focuses on machine learning methods for trajectory prediction in autonomous driving, with particular emphasis on how models and learned representations transfer across many real-world datasets. The aim is to improve understanding of cross-dataset generalization, transferability, and robustness in data-driven prediction systems through development, training, and evaluation across multiple datasets and experimental settings.