MediaPipeFaceMeshVisualize - ComfyUI Built-in Node Documentation

Overview

Draws face landmark points and connection lines (a face mesh) on top of an input image. This node uses the landmark data produced by a face detection node to visualize the detected facial features, such as the eyes, nose, mouth, and face outline.

Inputs

Parameter	Description	Data Type	Required	Range
`face_landmarks`	The face landmark data from a detection node.	FACE_LANDMARKS	Yes
`image`	The image to draw the mesh on. If not connected, a black canvas of the same size as the detection result will be used.	IMAGE	No
`connections`	Determines which parts of the face mesh to draw. `"all"` draws the full mesh (oval, eyes, brows, lips, irises, nose). `"fill"` draws a solid polygon of the face oval (silhouette mask). `"custom"` lets you toggle each feature individually. (default: `"all"`)	COMBO	Yes	`"all"` `"fill"` `"custom"`
`color`	The color of the mesh lines and points. (default: `#00ff00`)	COLOR	Yes
`thickness`	The thickness of the mesh lines in pixels. Setting this to 0 disables line drawing. (default: 1)	INT	Yes	0 to 8
`point_size`	The radius of the landmark dots in pixels. Setting this to 0 disables dot drawing. (default: 2)	INT	Yes	0 to 16

Note on connections parameter: When "custom" is selected, additional boolean inputs appear for each facial feature (e.g., face_oval, lips, left_eye, right_eye, left_eyebrow, right_eyebrow, left_iris, right_iris, nose, tesselation). Only the features you enable will be drawn.

Outputs

Output Name	Description	Data Type
`IMAGE`	The input image with the face landmarks mesh drawn on it.	IMAGE

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): fb5437d73378b0c8daa68669c2e19058ccb7133ed68fc51c8d4c5bab8662f243

MediaPipeFaceMask - ComfyUI Built-in Node Documentation

RTDETR_detect - ComfyUI Built-in Node Documentation

​Overview

​Inputs

​Outputs

Overview

Inputs

Outputs