Add pipeline tag and improve metadata

Hi! I'm Niels from the Hugging Face community team.

This PR improves the model card metadata by:
1. Adding the `image-to-image` pipeline tag to help users discover the model when filtering by task.
2. Specifying `pytorch` as the `library_name`.
3. Linking the model card to the associated paper on the Hugging Face Hub: [When Less Is More: Simplicity Beats Complexity for Physics-Constrained InSAR Phase Unwrapping](https://huggingface.co/papers/2605.00896).

I've also kept the usage snippets and architecture details found in your repository.

README.md

@@ -1,14 +1,15 @@
 ---
 license: cc-by-4.0
+library_name: pytorch
+pipeline_tag: image-to-image
 tags:
-  - InSAR
-  - phase-unwrapping
-  - remote-sensing
-  - geophysics
-  - U-Net
-  - PyTorch
-  - ML4RS
-  - ICLR2026
+- InSAR
+- phase-unwrapping
+- remote-sensing
+- geophysics
+- U-Net
+- ML4RS
+- ICLR2026
 ---
 
 # 🛰️ InSAR Phase Unwrapping - Pretrained Models
@@ -19,7 +20,7 @@ Pretrained weights for all four U-Net architectural variants from:
 > Prabhjot Singh, Manmeet Singh
 > **Oral Presentation - ML4RS @ ICLR 2026**
 
-📄 [Paper (OpenReview)](https://openreview.net/forum?id=liJldeR5ZX) | 💻 [Code (GitHub)](https://github.com/prabhjotschugh/When-Less-is-More-InSAR-Phase-Unwrapping)
+📄 [Paper (Hugging Face)](https://huggingface.co/papers/2605.00896) | 📄 [Paper (OpenReview)](https://openreview.net/forum?id=liJldeR5ZX) | 💻 [Code (GitHub)](https://github.com/prabhjotschugh/When-Less-is-More-InSAR-Phase-Unwrapping)
 
 ---