Privacy-Preserving Action Recognition: Taxonomy, Methods, and Privacy-Utility Trade-offs

2026-08-05

AI Summary

This study addresses the challenge of balancing privacy preservation and recognition utility in video-based action recognition by systematically reviewing 32 relevant works and organizing them into a five-category technical framework encompassing adversarial learning, skeleton-based modeling, cryptographic approaches, differential privacy, and hybrid methods. The work innovatively introduces a PRISMA-guided systematic review methodology, a formal threat model, a unified evaluation protocol termed PPAR, and a deployment-oriented standardization roadmap. It further proposes a two-dimensional privacy space taxonomy to clarify the trade-offs among privacy, utility, and efficiency. Findings reveal that only 10% of existing studies adopt formal privacy definitions; skeleton-based methods achieve up to 85% accuracy yet discard appearance information, while adversarial approaches maintain nearly 80% recognition utility under moderate privacy guarantees.

Abstract

Video surveillance in public safety, healthcare, and smart environments has made continuous human monitoring routine, raising real risks to personal identity and appearance. Privacy-preserving action recognition (PPAR) tackles the tension between the utility of video understanding and this exposure, and has drawn fast-growing interest. However, existing surveys remain narrow. Most catalog a single mechanism family, predate recent adversarial and hybrid work, or barely address evaluation. The result is a fragmented literature with incompatible threat models, inconsistent metrics, and no shared evaluation standard. We address this with a PRISMA-guided review of 32 peer-reviewed papers (2018--2026) drawn from 885 screened records. Methods sort into five families, namely adversarial learning (52%), skeleton-based (20%), cryptographic (12%), differential privacy (8%), and hybrid (8%), each with distinct privacy, utility, and efficiency trade-offs. Evaluation is the weak point. Only 10% of papers adopt a formal privacy definition, 65% rely on ad-hoc metrics, and 40% report an inconsistently defined cMAP. The trade-offs are steep. Skeleton methods reach about 85% accuracy but drop appearance, adversarial methods hold near 80% utility at moderate privacy (cMAP 0.9 to 0.3--0.5), and differential privacy often falls below 70%. Harder conditions stay under-tested, with fewer than 15% of papers checking cross-dataset generalization, under 10% testing adaptive attackers, and real-time edge deployment nearly untouched. We contribute a two-dimensional privacy-space taxonomy, a formal threat model, a comparative trade-off analysis, the PPAR Unified Evaluation Protocol, and a roadmap centered on benchmark standardization. With this grounding, we argue PPAR can move from prototypes toward deployment, with lessons extending to face recognition and medical imaging.

Problem

• Privacy-Preserving Action Recognition
• evaluation standardization
• privacy-utility trade-offs
• threat models
• benchmarking

Innovation

• privacy-preserving action recognition
• privacy-utility trade-off
• unified evaluation protocol
• adversarial learning
• differential privacy