Pose estimation

Thu Feb 03 2022

Pose estimation refers to detecting keypoint locations that describe the overall shape (skeleton) of an object. Here, we provide human pose estimation to jointly detect human body, hand, facial, and foot keypoints in a single image. Two popular pose estimator are represented: OpenPose and MediaPipe. OpenPose is a multi-person bottom-up method while MediaPipe is top-down and in its standard version, it is single person. Different set of keypoints is extracted using each method that you may find the details in the corresponding white paper. Human pose estimation have wide applications such as motion analysis, action recognition and more. To customize this service on your specific problem, please fill “Request for Customization” form. We have already make a sample yet useful service using pose estimation, i.e. “Corrective Excursive”. This service is also available with a free public demo in your user panel under Advanced Services category. Generic saiwa human pose estimation service has key features and tools including:

Supporting both bottom-up and top-down approaches
Fast and robust
Supporting both single-person and multi-person methods
User defined visualization threshold
Exporting and archiving the results in user cloud space or locally
Preview and download the resulting images or keypoint locationsHuman pose estimation have many applications like: 1. Activity recognition a. Elderly fall detection b. Autonomously corrective exercise c. Health condition estimation 2. Augmented reality Motion capture 3. Robotics Training with human pose tracking 4. Motion tracking 5. Sign language recognition 6. Security and surveillance