Improve model card metadata and content

README.md

@@ -1,17 +1,19 @@
 ---
-{}
+library_name: transformers
+pipeline_tag: image-feature-extraction
+license: other
 ---
 
-
 # StyleID — Stylization-Agnostic Identity Encoder
+
 [![arXiv](https://img.shields.io/badge/arXiv-2604.21689-b31b1b.svg)](https://arxiv.org/abs/2604.21689)
 [![Project Page](https://img.shields.io/badge/Project-Page-blue)](https://kwanyun.github.io/StyleID_page/)
+[![GitHub](https://img.shields.io/badge/GitHub-Code-black?logo=github)](https://github.com/kwanyun/StyleID)
 
+StyleID is a CLIP-based image encoder trained to produce identity embeddings that are robust to stylization (e.g., cartoons, sketches, and paintings). It is designed to maintain identity recognition even when artistic styles cause significant domain shifts in texture, color, or geometry.
 
-StyleID is a CLIP-based image encoder trained to produce identity embeddings that are robust to stylization.  
 It can be used for identity similarity, retrieval, evaluation, and conditioning in generative models.
 
-
 <img src="https://cdn-uploads.huggingface.co/production/uploads/639d445524af4747d8d2af52/1pTEZ88YvwnbDPlV_UqpM.jpeg" width="700">
 
 ---
@@ -19,28 +21,57 @@ It can be used for identity similarity, retrieval, evaluation, and conditioning
 ## Installation
 
 ```bash
-pip install transformers pillow
+pip install transformers pillow torch
 ```
 
 ## Usage
-#### Do not use for multiple faces or faces too small to recognize.
 
-```
+StyleID is best suited for single-face images. A rough center crop near the face is recommended for better performance. It is not intended for use with multiple faces or faces too small to recognize.
+
+```python
 import torch
 from transformers import CLIPModel, CLIPProcessor
 from PIL import Image
 
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
+# Load model and processor
 model = CLIPModel.from_pretrained("kwanY/styleid").to(device)
 processor = CLIPProcessor.from_pretrained("kwanY/styleid")
 
+# Prepare image
+img_path = "example.jpg"
 img = Image.open(img_path).convert("RGB")
 inputs = processor(images=img, return_tensors="pt").to(device)
 
+# Extract features
 with torch.no_grad():
     emb = model.get_image_features(**inputs)
     emb = emb / emb.norm(dim=-1, keepdim=True)  # optional but recommended
 ```
 
-#### Open for non-commercial research. Do not use FFHQ for biometric human recognition
+## Intended Uses
+
+- Identity similarity comparison
+- Image retrieval
+- Stylized identity evaluation
+- Identity-aware conditioning for generative models
+- Research on face recognition under domain shift and stylization
+
+## License and Usage Notice
+
+- StyleID is released for non-commercial research use.
+- Do not use FFHQ-derived data for biometric human recognition.
+
+## Citation
+
+If you find this work useful, please cite the paper:
+
+```bibtex
+@article{yun2026styleid,
+  title={StyleID: A Perception-Aware Dataset and Metric for Stylization-Agnostic Facial Identity Recognition},
+  author={Yun, Kwan and Lee, Changmin and Jeong, Ayeong and Kim, Youngseo and Lee, Seungmi and Noh, Junyong},
+  journal={arXiv preprint arXiv:2604.21689},
+  year={2026}
+}
+```