Update model card: add paper link, pipeline tag, and fix license

Hi! I'm Niels from the Hugging Face community science team. I'm opening this PR to improve the model card for LPNSR.

Summary of changes:
- Added the `image-to-image` pipeline tag to improve discoverability.
- Updated the license to `mit` based on the official GitHub repository.
- Added a link to the paper: [LPNSR: Prior-Enhanced Diffusion Image Super-Resolution via LR-Guided Noise Prediction](https://huggingface.co/papers/2603.21045).
- Added a link to the official GitHub repository.
- Included basic usage instructions for inference and the local Gradio demo.

README.md

@@ -1,14 +1,68 @@
----
-license: apache-2.0
-library_name: pytorch
-tags:
-- pytorch
-- computer-vision
-- image-super-resolution
----
-
-# LPNSR
-
-These are the official model weights (`.pth` files) for the [LPNSR project](https://github.com/Faze-Hsw/LPNSR) hosted on GitHub.
-
-To use these weights, clone the repository above and place the `.pth` files as specified in the repo instructions.
+---
+library_name: pytorch
+license: mit
+pipeline_tag: image-to-image
+tags:
+- pytorch
+- computer-vision
+- image-super-resolution
+- diffusion
+---
+
+# LPNSR: Prior-Enhanced Diffusion Image Super-Resolution via LR-Guided Noise Prediction
+
+This repository contains the official model weights for **LPNSR**, a prior-enhanced efficient diffusion framework for image super-resolution (SR).
+
+- **Paper:** [LPNSR: Prior-Enhanced Diffusion Image Super-Resolution via LR-Guided Noise Prediction](https://huggingface.co/papers/2603.21045)
+- **GitHub Repository:** [Faze-Hsw/LPNSR](https://github.com/Faze-Hsw/LPNSR)
+
+## Introduction
+
+LPNSR addresses the efficiency-quality trade-off in diffusion-based SR. While state-of-the-art frameworks like ResShift achieve efficient 4-step inference, they can suffer from performance degradation due to unconstrained random noise. LPNSR addresses this by:
+- Replacing random Gaussian noise with an **LR-guided multi-input-aware noise predictor**, embedding structural priors into the reverse process.
+- Mitigating initialization bias using a **high-quality pre-upsampling network** to optimize the diffusion starting point.
+- Maintaining a compact 4-step sampling trajectory for high-quality, real-world super-resolution.
+
+## Features
+
+- **Efficient Sampling**: Only 4 sampling steps required for high-quality super-resolution.
+- **Noise Predictor**: Learns to predict optimal noise maps for partial diffusion initialization.
+- **Real-world SR**: Designed to handle complex real-world degradations.
+- **SwinIR Integration**: Optional SwinIR refinement for enhanced details.
+
+## Quick Start
+
+To use these weights, clone the [official repository](https://github.com/Faze-Hsw/LPNSR) and follow the environment setup instructions.
+
+### Inference
+
+Once the environment is set up and weights are placed in the `pretrained/` folder, run:
+
+```bash
+python LPNSR/inference.py -i [image folder/image path] -o [output folder]
+```
+
+### Online Demo
+
+You can also launch a local Gradio web interface:
+
+```bash
+python LPNSR/app.py
+```
+
+## Citation
+
+If you find this work useful, please cite:
+
+```bibtex
+@article{lpnsr2026,
+  title={LPNSR: Prior-Enhanced Diffusion Image Super-Resolution via LR-Guided Noise Prediction},
+  author={[]},
+  journal={arXiv preprint arXiv:2603.21045},
+  year={2026}
+}
+```
+
+## Acknowledgement
+
+This project is based on [ResShift](https://github.com/zsyOAOA/ResShift), [BasicSR](https://github.com/XPixelGroup/BasicSR), [SwinIR](https://github.com/JingyunLiang/SwinIR), and [Real-ESRGAN](https://github.com/xinntao/Real-ESRGAN).