BEV-Net: Assessing Social Distancing Compliance by Joint People Localization and Geometric Reasoning

TL;DR

This paper introduces BEV-Net, a novel multi-branch neural network that localizes people and assesses social distancing compliance in crowded scenes using bird's eye view geometry, aiding public health monitoring.

Contribution

The work presents a new dataset, evaluation measures, and a multi-branch network architecture that jointly localizes individuals and detects social distancing violations in complex scenes.

Findings

BEV-Net outperforms existing baselines in crowded scene analysis.

The proposed system accurately localizes individuals in world coordinates.

Publicly available datasets and code facilitate further research.

Abstract

Social distancing, an essential public health measure to limit the spread of contagious diseases, has gained significant attention since the outbreak of the COVID-19 pandemic. In this work, the problem of visual social distancing compliance assessment in busy public areas, with wide field-of-view cameras, is considered. A dataset of crowd scenes with people annotations under a bird's eye view (BEV) and ground truth for metric distances is introduced, and several measures for the evaluation of social distance detection systems are proposed. A multi-branch network, BEV-Net, is proposed to localize individuals in world coordinates and identify high-risk regions where social distancing is violated. BEV-Net combines detection of head and feet locations, camera pose estimation, a differentiable homography module to map image into BEV coordinates, and geometric reasoning to produce a BEV map of the people locations in the scene. Experiments on complex crowded scenes demonstrate the power of the approach and show superior performance over baselines derived from methods in the literature. Applications of interest for public health decision makers are finally discussed. Datasets, code and pretrained models are publicly available at GitHub.