back

by jtwaleson·3y ago·view on hn ↗
The article states it could be really improved by using the Iris model, but I believe the MediaPipe Iris model is embedded in the FaceMesh model. Please correct me if I'm wrong.
1 comments
Iris is a separate model from face mesh, but it relies on face mesh for part of the pipeline

> The first step in the pipeline leverages MediaPipe Face Mesh, which generates a mesh of the approximate face geometry. From this mesh, we isolate the eye region in the original image for use in the subsequent iris tracking step. …

> Face Landmark Model

> The face landmark model is the same as in MediaPipe Face Mesh. You can also find more details in this paper.

> Iris Landmark Model

> The iris model takes an image patch of the eye region and estimates both the eye landmarks (along the eyelid) and iris landmarks (along this iris contour)…

https://google.github.io/mediapipe/solutions/iris