Pose On The Go 1.0.0

Pose-on-the-Go presents a comprehensive full-body pose estimation system that utilizes sensors already integrated into modern smartphones. Unlike previous approaches that depend on external or wearable sensors, this system employs extensive sensor fusion by leveraging the device’s front and rear cameras, user-facing depth camera, touchscreen, and inertial measurement unit (IMU). Despite these advancements, some critical data about the user's body—such as joint angles like the elbow—is inherently absent. To address this limitation, the system applies inverse kinematics techniques to estimate and animate plausible body poses. The methodology was published at ACM CHI 2021. Further details can be found at: https://karan-ahuja.com/poseotg.html.