| --- |
| license: mit |
| size_categories: |
| - 1K<n<10K |
| --- |
| |
| * Hair Area Segmentation Result (from face images, **Soft Label (for K.D.)**) using [FaceXFormer Pre-trained Model](https://kartik-3004.github.io/facexformer/). |
| * **Including 4,930 Segmentation Mask Images** |
| * [Original Face Images Dataset](https://huggingface.co/datasets/daebakgazua/250526_OhLoRA_filtered_images) |
| * for [Oh-LoRA v4 Project](https://github.com/WannaBeSuperteur/AI_Projects/tree/main/2025_06_24_OhLoRA_v4), |
| * Did **Response-Based Knowledge Distillation** for creating [Light-weighted Segmentation Model](https://huggingface.co/daebakgazua/250624_OhLoRA_Hair_Segmentation) for Oh-LoRA v4, with **FaceXFormer Pre-trained Model** as teacher |
| * [Detailed Info](https://github.com/WannaBeSuperteur/AI_Projects/tree/main/2025_06_24_OhLoRA_v4/segmentation) |
