mmCounter: Static People Counting in Dense Indoor Scenarios Using mmWave Radar

2025-12-11

AI Summary

Millimeter-wave (mmWave) radar struggles to accurately count static individuals in confined indoor spaces—particularly at densities ≤3 persons/m²—due to limited spatial resolution and the reliance of conventional approaches on motion-based detection. To address this, we propose a novel, multi-stage framework for low-frequency physiological signal separation and spatial source mapping that operates without prior assumptions about the number of subjects. Our method is the first to precisely decouple and localize individual respiratory and subtle torso-movement signals (<1 Hz) from radar returns. It integrates ultra-low-frequency feature extraction, adaptive noise suppression, and physiology-informed unsupervised clustering for contactless counting. Evaluated in familiar environments, it achieves an F1-score of 87% (MAE = 0.6); in unseen environments, F1 drops to 60% (MAE = 1.1). The system supports dense counting of up to seven individuals within a 3 m² area. This work overcomes a fundamental limitation in radar-based sensing of static populations and establishes a new paradigm for non-contact, high-density occupancy monitoring.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Humans and AI: Crowd Sourcing and Human Computation

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

mmWave radars struggle to detect or count individuals in dense, static (non-moving) groups due to limitations in spatial resolution and reliance on movement for detection. We present mmCounter, which accurately counts static people in dense indoor spaces (up to three people per square meter). mmCounter achieves this by extracting ultra-low frequency (< 1 Hz) signals, primarily from breathing and micro-scale body movements such as slight torso shifts, and applying novel signal processing techniques to differentiate these subtle signals from background noise and nearby static objects. Our problem differs significantly from existing studies on breathing rate estimation, which assume the number of people is known a priori. In contrast, mmCounter utilizes a novel multi-stage signal processing pipeline to extract relevant low-frequency sources along with their spatial information and map these sources to individual people, enabling accurate counting. Extensive evaluations in various environments demonstrate that mmCounter delivers an 87% average F1 score and 0.6 mean absolute error in familiar environments, and a 60% average F1 score and 1.1 mean absolute error in previously untested environments. It can count up to seven individuals in a three square meter space, such that there is no side-by-side spacing and only a one-meter front-to-back distance.

Problem

• Count static people in dense indoor groups using mmWave radar
• Extract ultra-low frequency signals from breathing and micro-movements
• Differentiate subtle human signals from noise and static objects

Innovation

• Extracts ultra-low frequency signals from breathing and micro-movements
• Applies novel multi-stage signal processing to differentiate signals
• Maps low-frequency sources with spatial information to individuals