search

wywu / LAB

[CVPR 2018] Look at Boundary: A Boundary-Aware Face Alignment Algorithm
https://wywu.github.io/projects/LAB/LAB.html
Other
1.01k stars 273 forks source link

Train/val/test split on 300W dataset? #10

Closed YuqiHUO closed 5 years ago

YuqiHUO commented 5 years ago

Thank you for your excellent work. Could you provide the split of train/val/test set on 300W dataset? In the paper you said all the data are used in train and other are used in test, you didn't mention validation set.

wywu commented 5 years ago

Hi,

We split the 300W dataset as most of the previous literature used to do.

  1. training set (3148 images) consisting of AFW (337), LFPW training set (811) and HELEN training set (2000)
  2. validation set: (a) the common subset (554) of HELEN testing set (330) and LFPW testing set (224) (b) the challenging subset (135) named IBUG (c) the full set (689 images) consisting of both the common (554) and challenging subsets (135).
  3. testing set: the private test-set used for the 300W competition which contains 600 images (300 indoor and 300 outdoor)

Best, Wayne

YuqiHUO commented 5 years ago

Thank you for your reply, so the common subset and challenging subset are both validation set, I know.