YOLOv8WithOpenCVForUnityExample icon indicating copy to clipboard operation
YOLOv8WithOpenCVForUnityExample copied to clipboard
verified publisher icon EnoxSoftware

Any way to get facial landmark detection?

Open thePostFuturist opened this issue 1 year ago • 2 comments

YOLO is light years ahead of Mediapipe -- extremely robust. This model does track the nose, eyes, and ears -- is there another model or a way to get it to track more face landmarks?

thePostFuturist avatar Feb 28 '24 04:02 thePostFuturist

i still recommend using dlib68 for facial landmarks

yosun avatar Mar 09 '24 11:03 yosun

dlib68 is very mid, doesnt come close to ARKit's tracking

thePostFuturist avatar Jun 17 '24 05:06 thePostFuturist