Update README.md

README.md

@@ -1,3 +1,86 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+---
+
+---
+
+## 🎨 PosterCraft: Rethinking High-Quality Aesthetic Poster Generation in a Unified Framework
+
+[![demo](https://img.shields.io/badge/%F0%9F%96%8C%EF%B8%8F-Demo-000000?style=flat-square\&logo=vercel)](https://ephemeral182.github.io/PosterCraft/)
+[![arxiv](https://img.shields.io/badge/arXiv-Preprint-b31b1b?style=flat-square)](https://arxiv.org/abs/xxxx.xxxxx)
+[![github](https://img.shields.io/badge/%F0%9F%92%BB-Code-181717?style=flat-square\&logo=github)](https://github.com/ephemeral182/PosterCraft)
+[![video](https://img.shields.io/badge/%F0%9F%8E%A5-Video-ff0000?style=flat-square\&logo=youtube)](https://www.youtube.com/watch?v=XXXXXX)
+[![HF Spaces](https://img.shields.io/badge/%F0%9F%A4%96-HuggingFace%20Demo-orange?style=flat-square\&logo=huggingface)](https://huggingface.co/spaces/ephemeral182/postercraft-demo)
+
+---
+
+### ✨ Overview
+
+**PosterCraft-v1\_RL** is a high-quality aesthetic poster generation model trained with reinforcement learning on multi-stage synthetic and curated datasets. It takes a structured prompt and produces posters with strong visual composition, artistic balance, and accurate typography rendering.
+
+This version is inference-only and ideal for **design inspiration**, **AI-driven visual layout generation**, and **creative content production**.
+
+> 🔗 **Live Demo**: [ephemeral182.github.io/PosterCraft](https://ephemeral182.github.io/PosterCraft/)
+> 🧠 **Model Type**: Custom Diffusion / Poster Generation
+> 🔁 **Inference Only** (weights available, training pipeline not included)
+
+---
+
+### 🖼️ Features
+
+* ✅ **Text-to-Poster generation** with multi-element structured layout
+* 🎯 **Fine-tuned with RL** using aesthetic feedback and layout score
+* ✍️ **Typographic region control** using Gemini-generated masks
+* 🧠 **Multi-modal reward design** (text, layout, mask)
+* ⚡ Fast inference with Hugging Face Spaces (check demo)
+
+---
+
+### 🚀 How to Use
+
+```python
+from diffusers import StableDiffusionPipeline
+
+pipe = StableDiffusionPipeline.from_pretrained("ephemeral182/PosterCraft-v1_RL").to("cuda")
+prompt = "A cinematic poster for 'The Future is Now': futuristic city, neon glow, lone figure, dramatic skyline"
+image = pipe(prompt).images[0]
+image.save("poster.png")
+```
+
+Alternatively, try the [Hugging Face Web Demo](https://huggingface.co/spaces/ephemeral182/postercraft-demo) or use the [official site](https://ephemeral182.github.io/PosterCraft/) for richer UI support and mask upload functionality.
+
+---
+
+### 📦 Resources
+
+| Resource         | Link                                                                               |
+| ---------------- | ---------------------------------------------------------------------------------- |
+| 🧪 Paper (arXiv) | [arxiv.org/abs/xxxx.xxxxx](https://arxiv.org/abs/xxxx.xxxxx)                       |
+| 🧠 Code (GitHub) | [github.com/ephemeral182/PosterCraft](https://github.com/ephemeral182/PosterCraft) |
+| 🎬 Video Intro   | [YouTube Walkthrough](https://www.youtube.com/watch?v=XXXXXX)                      |
+| 🌐 Website       | [PosterCraft WebApp](https://ephemeral182.github.io/PosterCraft/)                  |
+
+---
+
+### 📌 License
+
+This model is released under the [Apache 2.0 license](https://www.apache.org/licenses/LICENSE-2.0).
+
+---
+
+### 🧩 Citation
+
+```
+@misc{chen2025postercraft,
+  title={PosterCraft: Rethinking High-Quality Aesthetic Poster Generation in a Unified Framework},
+  author={Chen, Sixiang and others},
+  year={2025},
+  eprint={xxxx.xxxxx},
+  archivePrefix={arXiv},
+  primaryClass={cs.CV}
+}
+```
+
+---
+
+---