Explore datasets and labels
Alpha images perfectly separate the character from the background.
Applications:
The “surface normal” is a vector that points directly away from the surface, at a
right angle to it, at that point.
Applications:
Z-Depth in meters from the camera at every point.
Applications:
We provide 43 classes of subsegmentation on the face, and 50 classes of the body,
consistent with industry best practices.
Applications:
These basic facial landmarks replicate iBug 68 locations.
Applications:
MediaPipe Face Mesh is a facial landmark detection model developed by Google. It
can detect 468 3D facial landmarks on a human face in real-time.
Applications:
Synthesis AI’s 5,000 dense landmarks are particularly useful for training models
that require a highly detailed understanding of the human face; e.g., in ID verification and
security applications.
Applications:
The centroid of the eye volume, to the location of the pupil in world
coordinates, valuable for attention monitoring in ADAS.
Applications:

Our basic facial landmarks are similar, but not identical, toiBug 68 locations.
Applications:

The centroid of the eye volume to the location of the pupil in world coordinates, valuable for attention monitoring.
Applications:

We provide a large library of typical activities showcasing distracted driving.
Applications:

Synthesis AI’s 5,000 dense landmarks are particularly useful for training models that require a highly detailed understanding of the human face; e.g., in ID verification and security applications.
Applications:

MediaPipe Face Mesh is a solution that estimates 468 3D face landmarks.
Applications:

We provide our dense landmark model in the .obj format, with over 4,000 3D landmarks.
Applications:

Alpha perfectly separates the character from the background.
Applications:

The “surface normal” is a vector that points directly away from the surface, at a right angle to it, at that point.
Applications:




