The MediaPipe Holistic Landmarker task lets you combine components of the
pose, face, and
hand landmarkers to create a complete landmarker for
the human body. You can use this task to analyze full-body gestures, poses, and
actions. This task uses a machine learning (ML) model on a continuous stream of
images. The task outputs a total of 543 landmarks (33 pose landmarks, 468 face
landmarks, and 21 hand landmarks per hand) in real-time.
An upgraded version of this MediaPipe Solution is coming soon! The MediaPipe Legacy
Solution for this task is available on
GitHub.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2024-04-24 UTC."],[],[]]
