Pocket Pose is a deep learning framework that involves predicting the position and orientation of body parts in images or videos.This framework employs precise deep learning models to simultaneously detect body, face, hands, and foot keypoints in real-time. It stands out for its versatility and efficiency, capable of processing multiple people in complex poses and varied environments. Its architecture effectively captures spatial relationships between keypoints through carefully designed novel architectures that improves the accuracy and robustness of pose detection.