Detect eye blink?

Are the ret landmarks suitable to compute EAR？
more see https://pyimagesearch.com/2017/04/24/eye-blink-detection-opencv-python-dlib/
def eye_aspect_ratio(eye):
# compute the euclidean distances between the two sets of
# vertical eye landmarks (x, y)-coordinates
A = dist.euclidean(eye[1], eye[5])
B = dist.euclidean(eye[2], eye[4])

# compute the euclidean distance between the horizontal
# eye landmark (x, y)-coordinates
C = dist.euclidean(eye[0], eye[3])

# compute the eye aspect ratio
ear = (A + B) / (2.0 * C)

# return the eye aspect ratio
return ear

The issue asks if the library's facial landmarks (likely from the 'ret' model) are suitable for computing the Eye Aspect Ratio (EAR) for blink detection. First, examine the landmark indices and coordinate output in the face-alignment library's documentation or source code (e.g., in face_alignment/api.py or face_alignment/models.py). Compare the landmark ordering to the dlib model used in the linked article. Then, write a small script to extract eye landmarks from a test image and compute EAR to verify suitability. The goal is to confirm whether the existing landmarks can be directly used for blink detection or if preprocessing is needed.

Written by the indexing model from the issue text.