Find Faces in Image (vuo.image.findFaces)

Tries to find human faces in an image, including eye, nose, and mouth positions.

• Minimum Confidence — The threshold for detecting a face, ranging from 0 (detect more faces, with more false positives) to 1 (detect fewer faces, with fewer false positives).
• Quality — Controls the tradeoff between speed and thoroughness, ranging from 0 (fast, but detects only faces occupying a large portion of the image) to 1 (slower, but detects smaller faces).
• Rectangles — The location and size of all detected faces, in Vuo Coordinates.
• Landmarks — The location of each feature of the face (eyes, nose, mouth), in Vuo Coordinates. Use Get Face Landmarks to get individual values.

Since finding faces is a computationally intensive operation, you’ll probably want to set the Event Throttling mode of the trigger port driving this node to “Drop Events”.

Keywords: cheeks, chin, detection, eyebrows, features, forehead, head, human, left eye, lips, motion, mouth, nose, person, premium, pro, right eye, search, tracking, vision

Example compositions:

• Mark Faces In Image: Draws a rectangle over each face detected in the dropped image.
• Highlight Eyes in Video: Darkens the background surrounding the human eyes detected in the camera image.
• Obscure Faces In Video: Pixellates each human face in the camera's view.

This is a pro node. If you share this composition with other Vuo users, they'll need to have access to pro nodes in order to run it.