Privacy-Preserving Sensing in Consumer Security Cameras: A Survey of mmWave Radar, PIR, Vision, and Hybrid Sensor Fusion

Consumer security cameras have proliferated significantly in the post-COVID era, raising substantial concerns around privacy and regulatory compliance. Traditional vision-based systems offer high spatial and semantic resolution but at the cost of constant visual surveillance, raising alarms in GDPR-compliant regions. In contrast, modalities such as mmWave radar and passive infrared (PIR) sensors offer inherently more privacy-conscious alternatives. However, a lack of unified comparative analysis across these sensing modalities—particularly in terms of privacy exposure, power consumption, environmental robustness, and regulatory constraints—limits the development of principled designs for next-generation consumer security systems. This survey synthesizes recent advances across radar, PIR, vision, and hybrid sensor systems. This survey contributes a taxonomy of privacy-preserving sensing, a quantitative cross-modal comparison, an analysis of deployment trade-offs, and a discussion of regulatory implications. The key insight is that hybrid sensor fusion architectures—combining low-resolution radar or PIR sensors with selective vision or audio—are emerging as a dominant paradigm. These systems can preserve privacy while maintaining acceptable detection fidelity, especially when paired with on-device inference.