Mind the Gap: Mapping Wearer–Bystander Privacy Tensions and Context-Adaptive Pathways for Camera Glasses

Paper Title

Mind the Gap: Mapping Wearer–Bystander Privacy Tensions and Context-Adaptive Pathways for Camera Glasses

Publication Info

• Topic area: Privacy challenges and solutions in wearable camera glasses.
• Keywords: Camera glasses, privacy, bystander protection, privacy-enhancing technologies, context-aware design, wearable devices, surveillance, user studies, multi-stakeholder analysis, ubiquitous sensing.

Background and Problem

• Problem / challenge: Camera glasses create privacy tensions between wearers and bystanders due to covert recording capabilities and inadequate notification mechanisms. Existing Privacy-Enhancing Technologies (PETs) fail to reconcile these conflicts effectively.
• Significance: Addressing these tensions is crucial for the social acceptance of camera glasses and for protecting individuals from unauthorized surveillance and privacy violations.
• Motivation and related work: Prior research has documented privacy concerns and proposed PETs, but gaps remain in systematically measuring expectation differences between wearers and bystanders, evaluating PET effectiveness, and understanding how context influences privacy dynamics.

Solution

• Proposed approach: A context-adaptive privacy framework that dynamically adjusts protection strategies based on environmental characteristics, stakeholder roles, and social norms.
• Novelty:
  1. Quantification of expectation-willingness gaps between wearers and bystanders across privacy dimensions and contexts.
  2. Identification of four fundamental trade-offs in PET design: visibility vs. disruption, empowerment vs. burden, protection vs. agency, and accountability vs. exposure.
  3. Proposal of context-adaptive pathways for privacy protection in public, semi-public, and sensitive environments.
• Procedure and key techniques:
  • Conducted a two-study investigation: a large-scale survey (N=525) to measure privacy expectation gaps and paired interviews (N=20) to evaluate 12 PETs.
  • Developed a diagnostic framework to analyze privacy mechanisms and proposed adaptive strategies for different contexts.

Results

• Concrete findings:
  • Bystanders demand stronger privacy protections (e.g., data sharing control, prior consent) than wearers are willing to provide, with gaps intensifying in sensitive contexts.
  • LED indicators, the most common notification mechanism, are deemed inadequate by both groups (only 41.3% of bystanders and 35.3% of wearers found them sufficient).
  • Defensive actions by bystanders (e.g., avoiding cameras, requesting data deletion) are common, with 65–90% indicating they would take action in sensitive contexts.
• Advantage over baselines:
  • Context-aware PETs (e.g., geofencing, face anonymization) outperform static mechanisms by addressing privacy needs dynamically, achieving higher ratings for privacy protection and usability.
  • Adaptive pathways resolve key trade-offs by tailoring protection strategies to specific environments, reducing conflicts between wearers and bystanders.
• Experiments / evaluation:
  • Study 1: Surveyed wearers and bystanders across six scenarios varying by physical setting and social relationship, revealing expectation-willingness gaps.
  • Study 2: Evaluated 12 PETs with HCI researchers and experienced users, identifying trade-offs and context-specific preferences.
• Limitations and future work:
  • Cultural specificity to China limits generalizability; cross-cultural studies are needed.
  • Scenario-based evaluations may not fully capture real-world complexities.
  • Future work should focus on lightweight context recognition systems, privacy-preserving negotiation protocols, and global regulatory frameworks.

Summary

This paper systematically investigates privacy tensions between wearers and bystanders of camera glasses, revealing significant expectation-willingness gaps and identifying four fundamental trade-offs in PET design. Through surveys and interviews, the authors propose a context-adaptive privacy framework that dynamically adjusts protection strategies based on environmental and social factors. The findings emphasize the need for adaptive, multi-stakeholder solutions to reconcile privacy conflicts, offering actionable insights for designing socially acceptable wearable technologies.