MediaPipe Solutions

Easy, ready-made, plug-and-play ML features.

Think of Solutions as prebuilt tools created by Google.

You simply import them and use them — no ML, no pipeline building, no graph editing.

✔ What you get

• Hands tracking
• Pose tracking
• Face Mesh
• Face Detection
• Object Detection
• Holistic (full body)
• Tasks (Vision, Audio, Text)

These Solutions:

• Already contain the ML models
• Already define the MediaPipe graph
• Already handle post-processing
• Already handle tracking, smoothing, 3D projection, etc.

Everything is ready to use, like calling a function.

✔ Example of using a Solution (Pose)

import mediapipe as mp
mp_pose = mp.solutions.pose

pose = mp_pose.Pose()

results = pose.process(rgb_image)