Member-only story

[Quick-ML] Introduction To Pose Estimation

Easy to digest information on Pose Estimation

2 min readAug 31, 2021

Pose Estimation is the process of deducing the location of the main joints of a body from provided digital assets such as images, videos or a sequence of images.
Pose estimation is a form of motion analysis that provides information on the location, pose and orientation of essential body parts in a human frame.

Pose estimation is made possible through several computer vision techniques. Forms of pose estimation are present in applications such as Action recognition, Human interactions, virtual reality assets and 3d graphics games, robotics and more.

Pose estimation solutions engineered as applications are applied within several industries. Although most of its applications are currently in research phases, Pose estimation solutions can generate movement dynamics for animated characters within video games or animated motion pictures through the analysis of a human actor's movement and bodily operations.
Within physiologically related fields, pose estimation is utilised for the in-depth understanding of motion analysis of a human. A better understanding of human body language or movement can aid treatments of patients suffering from locomotive dysfunction.
Pose estimation can be utilised in a variety of research fields an industry, for example, pose estimation can be used to generate animated characters for animated movies or video games by merely tracking the pose of a human actor to build a library of movement and actions for the animated character as opposed to traditional 3D graphics methods.
Another application can be found in psychology and physiology as the study of human body language and movement can be aided with an efficient method of human motion analysis provided by Pose estimation.

Previously, the conduction of pose estimation required the attachment of movement monitoring hardware on a target in a controlled environment.
Pose estimation and various forms of motion analysis operate with cameras and few monitoring devices. State of the art techniques that solve pose estimation requires only images or video sequences to conduct articulated pose estimation.