Hair Area Segmentation Result (from face images, Soft Label (for K.D.)) using FaceXFormer Pre-trained Model.
Including 4,930 Segmentation Mask Images
Original Face Images Dataset
for Oh-LoRA v4 Project,
Did Response-Based Knowledge Distillation for creating Light-weighted Segmentation Model for Oh-LoRA v4, with FaceXFormer Pre-trained Model as teacher
Detailed Info
