StickFigure enables privacy-friendly pose detection on tiny IoT cameras
StickFigure: Enabling Pose-Only IoT Cameras for Privacy-Preserving Edge Vision
Computer Vision and Pattern Recognition
Summary
Indoor cameras can help people by recognizing body movements but sending full video exposes private details. The authors created StickFigure, a system that keeps videos on the camera and only sends simple stick figure data showing poses. This approach protects privacy while still letting smart devices understand human actions. StickFigure runs efficiently on very small, low-power chips found in many IoT cameras and is more accurate and faster than previous solutions.
What this means in practice
- •For smart home developers: Integrate resource-efficient pose estimation on low-power IoT cameras to enable privacy-preserving activity-aware smart environments.
- •For assistive technology teams: Deploy on-device skeleton tracking in cameras to support real-time monitoring for assisted living without transmitting sensitive video data.
Authors
Aveline Xiaoyuan Han, Chenshu Wu
Abstract
Indoor IoT cameras can support assisted living and activity-aware spaces, but exporting raw RGB frames raises privacy concerns by exposing facial, clothing, and household details that body-motion tasks often do not require. We present StickFigure, a pose-only IoT camera system that, during real-time monitoring, keeps raw frames local and ephemeral while exporting only compact skeleton streams sufficient for downstream human activity recognition (HAR). While pose estimation has matured on GPUs, extracting accurate pose streams on resource-constrained IoT cameras remains difficult. To overcome these challenges and enable on-device pose streams on (almost) any IoT camera, we present two core designs in StickFigure: an activation-aware pose backbone that allocates channel capacity after spatial downsampling and a lightweight hierarchical MLP refiner that selectively corrects elbow and wrist coordinates using intermediate visual features and kinematic context. We fully implement StickFigure and evaluate it on public benchmarks and real-world deployment. Compared with state-of-the-art MCU pose models, StickFigure achieves the highest pose accuracy and downstream HAR utility on both NTU RGB+D and Penn Action, while using 96.5% fewer parameters and 99.3% fewer MACs per frame than RTMPose-S-based baselines, achieving 0.107 s p95 latency on a representative STM32H753 MCU. We believe StickFigure establishes a new operating point for privacy-aware edge vision on massive IoT cameras.