A Lightweight Crowd Model for Robot Social Navigation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Eskeri, Maryam Kazemi, Wiedemann, Thomas, Kyrki, Ville, Baumann, Dominik, Kucner, Tomasz Piotr
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909755178483712
author Eskeri, Maryam Kazemi
Wiedemann, Thomas
Kyrki, Ville
Baumann, Dominik
Kucner, Tomasz Piotr
author_facet Eskeri, Maryam Kazemi
Wiedemann, Thomas
Kyrki, Ville
Baumann, Dominik
Kucner, Tomasz Piotr
contents Robots operating in human-populated environments must navigate safely and efficiently while minimizing social disruption. Achieving this requires estimating crowd movement to avoid congested areas in real-time. Traditional microscopic models struggle to scale in dense crowds due to high computational cost, while existing macroscopic crowd prediction models tend to be either overly simplistic or computationally intensive. In this work, we propose a lightweight, real-time macroscopic crowd prediction model tailored for human motion, which balances prediction accuracy and computational efficiency. Our approach simplifies both spatial and temporal processing based on the inherent characteristics of pedestrian flow, enabling robust generalization without the overhead of complex architectures. We demonstrate a 3.6 times reduction in inference time, while improving prediction accuracy by 3.1 %. Integrated into a socially aware planning framework, the model enables efficient and socially compliant robot navigation in dynamic environments. This work highlights that efficient human crowd modeling enables robots to navigate dense environments without costly computations.
format Preprint
id arxiv_https___arxiv_org_abs_2508_19595
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Lightweight Crowd Model for Robot Social Navigation
Eskeri, Maryam Kazemi
Wiedemann, Thomas
Kyrki, Ville
Baumann, Dominik
Kucner, Tomasz Piotr
Robotics
Machine Learning
Robots operating in human-populated environments must navigate safely and efficiently while minimizing social disruption. Achieving this requires estimating crowd movement to avoid congested areas in real-time. Traditional microscopic models struggle to scale in dense crowds due to high computational cost, while existing macroscopic crowd prediction models tend to be either overly simplistic or computationally intensive. In this work, we propose a lightweight, real-time macroscopic crowd prediction model tailored for human motion, which balances prediction accuracy and computational efficiency. Our approach simplifies both spatial and temporal processing based on the inherent characteristics of pedestrian flow, enabling robust generalization without the overhead of complex architectures. We demonstrate a 3.6 times reduction in inference time, while improving prediction accuracy by 3.1 %. Integrated into a socially aware planning framework, the model enables efficient and socially compliant robot navigation in dynamic environments. This work highlights that efficient human crowd modeling enables robots to navigate dense environments without costly computations.
title A Lightweight Crowd Model for Robot Social Navigation
topic Robotics
Machine Learning
url https://arxiv.org/abs/2508.19595