Paper Image

Transferring vehicle movement patterns to new intersections

Published on:

2 April 2024

Primary Category:

Computer Vision and Pattern Recognition

Paper Authors:

Christoph Glasmacher,

Michael Schuldes,

Sleiman El Masri,

Lutz Eckstein


Key Details

Proposes methodology to model causal relations between driving scenario parameters

Uses Bayesian networks to analyze dependencies and decrease data needs

Transfers learned patterns to unseen intersections to generate scenarios

Evaluates on real-world trajectory dataset by comparing generated and recorded data

AI generated summary

Transferring vehicle movement patterns to new intersections

This paper proposes a methodology to analyze the relationships between parameters that define driving scenarios, using Bayesian networks. It aims to decrease the amount of required real-world data and enable the transfer of learned causal patterns to generate plausible scenarios for unobserved intersections. The method is evaluated by extracting scenarios from an intersection dataset, estimating movement patterns, transferring them to unseen intersections, and comparing against real-world trajectories.

Answers from this paper


No comments yet, be the first to start the conversation...

Sign up to comment on this paper

Sign Up