EffectivePose is an Android app that uses ML Kit to convert live video streams from a smartphone into accurate body pose estimations displayed in real-time. This project demonstrates the potential to make advanced computer vision technologies accessible to a wider audience and enhance user-device interactions without specialized tracking hardware.
I led the complete development lifecycle, focusing on integrating and optimizing machine learning models for high performance on various devices. Key enhancements included utilizing GPU processing reducing estimation times and improving frame rates, achieving up to 60 fps on high-end devices. This project underscores my expertise in mobile development, machine learning integration, and real-time processing optimization.