OpenFace – a state-of-the art tool intended for facial landmark detection, head pose estimation, facial action unit recognition, and eye-gaze estimation.
Other
6.71k
stars
1.82k
forks
source link
bboxdir text files does not work with `x_min y_min x_max y_max` format #1068
The wiki specifies -bboxdir to consist of text files with x_min y_min x_max y_max format. However, I noticed that it doesn't work (top image). If I input x_center y_center x_max y_max, I get a more accurate result from feature extraction (bottom image).
Is this the intended behavior? I notice that other face detector like MTCNN typically return bboxes in the same format (x_center y_center x_max y_max). Please advise.
Hello, Thank you for your work on OpenFace.
The wiki specifies
-bboxdir
to consist of text files withx_min y_min x_max y_max
format. However, I noticed that it doesn't work (top image). If I inputx_center y_center x_max y_max
, I get a more accurate result from feature extraction (bottom image)../bin/FaceLandmarkImg -fdir <src_dir> -bboxdir <bbox_dir> -wild
Is this the intended behavior? I notice that other face detector like MTCNN typically return bboxes in the same format (
x_center y_center x_max y_max
). Please advise.