Improving Crowd Density Estimation by Fusing Aerial Images and Radio Signals

Kai Wei Yang, Yen Yun Huang, Jen Wei Huang, Ya Rou Hsu, Chang Lin Wan, Hong Han Shuai, Li Chun Wang, Wen Huang Cheng

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

A recent line of research focuses on crowd density estimation from RGB images for a variety of applications, for example, surveillance and traffic flow control. The performance drops dramatically for low-quality images, such as occlusion, or poor light conditions. However, people are equipped with various wireless devices, allowing the received signals to be easily collected at the base station. As such, another line of research utilizes received signals for crowd counting. Nevertheless, received signals offer only information regarding the number of people, while an accurate density map cannot be derived. As unmanned aerial vehicles (UAVs) are now treated as flying base stations and equipped with cameras, we make the first attempt to leverage both RGB images and received signals for crowd density estimation on UAVs. Specifically, we propose a novel network to effectively fuse the RGB images and received signal strength (RSS) information. Moreover, we design a new loss function that considers the uncertainty from RSS and makes the prediction consistent with the received signals. Experimental results show that the proposed method successfully helps break the limit of traditional crowd density estimation methods and achieves state-of-the-art performance. The proposed dataset is released as a public download for future research.

Original languageEnglish
Article number84
JournalACM Transactions on Multimedia Computing, Communications and Applications
Volume18
Issue number3
DOIs
StatePublished - Aug 2022

Keywords

  • Crowd density estimation
  • Data fusion
  • Datasets
  • Unmanned aerial vehicles

Fingerprint

Dive into the research topics of 'Improving Crowd Density Estimation by Fusing Aerial Images and Radio Signals'. Together they form a unique fingerprint.

Cite this