EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Two-Stage Coarse-to-Fine Framework for Sparse Crowd Density Prediction in Digital Twin-Based Safety Monitoring

Younghwan Jeong iD, SoHyeon Kim iD, Jinyoung Lee iD, Donghoon Lee iD, Taemin Hwang iD, Won Gi Choi iD

Sensors · 27.06.2026 · 26 (13) · 4094

DOI: 10.3390/s26134094 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 27.06.2026. Verfasst von Younghwan Jeong und 5 weiteren Autor:innen. Beteiligte Einrichtung: Korea Electronics Technology Institute, Seongnam-si 13509, Gyeonggi-do, Republic of Korea. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Institute of Information & Communications Technology Planning & Evaluation, Korea Creative Content Agency, Institute of Information and Communications Technology Planning and Evaluation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Crowd-related disasters in dense public spaces unfold into hazardous situations within seconds, repeatedly demonstrating that reactive response alone is insufficient to minimize damage. This reality has intensified the need for monitoring systems that can proactively forecast congestion before it reaches a critical level. Digital twin platforms address this need by providing an operational substrate that represents crowd states on a unified bird’s-eye-view (BEV) grid, on which a predictive module can forecast where congestion will emerge. However, conventional AI-based single-stage dense prediction models are intrinsically ill-suited to this role: although crowd congestion is sparse in both space and time, these models apply uniform high-resolution computation across the entire BEV domain, wasting computation and biasing optimization toward dominant background regions. In this paper, we propose a two-stage coarse-to-fine framework that operates as the predictive module of the digital twin and explicitly exploits the spatio-temporal sparsity of crowd congestion. The first stage, CoarseSTFormer, performs efficient global screening on a low-resolution BEV input to coarsely identify a set of density-critical candidate regions. The second stage, SparseQueryDecoder, selectively reconstructs high-resolution responses only on the identified candidates, rather than uniformly upsampling the entire BEV grid. In simulation environments with up to 20,000 pedestrian agents, the proposed framework matches the strongest dense baseline in reconstruction quality while delivering the most balanced variance profile across grid scales. At inference, it further reduces GPU energy consumption by 1.9× to 5.0× and computational cost (FLOPs) by 3.8× to 54×, demonstrating its practicality as a resource-efficient predictive module that satisfies both accuracy and efficiency.

Schlagwörter

Stichwörter der Autor:innen

crowd density predictioncrowd safety monitoringdigital twinspatio-temporal sparsity

Autor:innen und Institutionen

  • Korea Electronics Technology Institute, Seongnam-si 13509, Gyeonggi-do, Republic of Korea.
  • Sonatus Korea, 660, Daewangpangyo-ro, Bundang-gu, Seongnam-si 13494, Gyeonggi-do, Republic of Korea.

ORCID-Profile: Younghwan Jeong, SoHyeon Kim, Jinyoung Lee, Donghoon Lee, Taemin Hwang, Won Gi Choi

Förderung

  • Institute of Information & Communications Technology Planning & Evaluation
  • Korea Creative Content Agency
  • Institute of Information and Communications Technology Planning and Evaluation
  • Korea Government through the Ministry of Science and ICT
  • Ministry of Culture, Sports and Tourism

Zitieren

APA

Jeong, Y., Kim, S., Lee, J., Lee, D., Hwang, T., & Choi, W. G. (2026). A Two-Stage Coarse-to-Fine Framework for Sparse Crowd Density Prediction in Digital Twin-Based Safety Monitoring. Sensors, 26(13), 4094. https://doi.org/10.3390/s26134094

Vancouver

Jeong Y, Kim S, Lee J, Lee D, Hwang T, Choi WG. A Two-Stage Coarse-to-Fine Framework for Sparse Crowd Density Prediction in Digital Twin-Based Safety Monitoring. Sensors. 2026;26(13):4094. doi:10.3390/s26134094

BibTeX
@article{Jeong2026,
  author = {Jeong, Younghwan and Kim, SoHyeon and Lee, Jinyoung and Lee, Donghoon and Hwang, Taemin and Choi, Won Gi},
  title = {A Two-Stage Coarse-to-Fine Framework for Sparse Crowd Density Prediction in Digital Twin-Based Safety Monitoring},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {13},
  pages = {4094},
  doi = {10.3390/s26134094}
}
Kostenlos starten →