Poseie

Poseie is a proposed open standard for encoding, transmitting, and interpreting human pose data across software applications and devices. Designed to support interoperable pose-based interactions in computer vision, animation, augmented reality, virtual reality, and robotics, Poseie seeks to provide a compact, expressive representation of body pose that can be shared between sensors, analyses, and rendering systems.

At its core, Poseie represents a pose as a set of joints with coordinates, confidence scores, and optional temporal information. The schema is modular, with a core pose model and optional extensions for three-dimensional orientation, limb connections, and scene context. The encoding supports both human-readable formats, such as JSON, for development, and a binary variant for bandwidth-constrained environments. Versioning and schemas are used to ensure backward compatibility.

Reference implementations and libraries associated with Poseie aim to integrate with outputs from established pose estimation systems and sensors. Use cases include animating avatars in games and film pipelines, enabling detailed sports analytics, guiding robotic controllers, and providing gesture-driven interfaces for accessibility and user interaction.

Privacy and ethics considerations are central to Poseie’s design, emphasizing data minimization, on-device processing, and user consent, particularly in consumer devices and public deployments. The standard has attracted interest from academic groups and industry pilots, but broad adoption remains limited due to competing formats, performance constraints, and fragmented toolchains across platforms. See also: Pose estimation, OpenPose, MediaPipe, motion capture, avatar streaming.