Improve model card and add metadata

README.md

@@ -1,4 +1,43 @@
-Official checkpoints and evaluation masks for <br>
-*Seeing Through Touch: Tactile-Driven Visual Localization of Material Regions (CVPR 2026)*.
+---
+pipeline_tag: image-segmentation
+---
 
-Please refer to [project page](https://mm.kaist.ac.kr/projects/SeeingThroughTouch/) for details.
+# Seeing Through Touch: Tactile-Driven Visual Localization of Material Regions
+
+Official checkpoints and evaluation masks for the paper **Seeing Through Touch: Tactile-Driven Visual Localization of Material Regions**, presented at CVPR 2026.
+
+[**Project Page**](https://mm.kaist.ac.kr/projects/SeeingThroughTouch/) | [**Paper**](https://huggingface.co/papers/2604.11579) | [**GitHub**](https://github.com/kaistmm/SeeingThroughTouch)
+
+## Introduction
+
+Tactile localization aims to identify image regions that share the same material properties as a tactile input. This model learns local visuo-tactile alignment via dense cross-modal feature interactions, producing tactile saliency maps for touch-conditioned material segmentation. 
+
+To overcome dataset constraints, the authors introduced in-the-wild multi-material scene images and a material-diversity pairing strategy to improve contextual localization and robustness.
+
+## Evaluation and Training
+
+For detailed instructions on environment setup, data preparation, evaluation, and training, please refer to the [official GitHub repository](https://github.com/kaistmm/SeeingThroughTouch).
+
+### Available Checkpoints
+
+The repository contains the following checkpoints:
+- `STT.pth`: Main SeeingThroughTouch (STT) model.
+- `STT-Local.pth`: Local-only variant.
+- `STT-Indomain.pth`: In-domain trained variant.
+
+## Citation
+
+If you find this work useful, please cite it as:
+
+```bibtex
+@inproceedings{kim2026seeingthroughtouch,
+  author    = {Seongyu Kim and Seungwoo Lee and Hyeonggon Ryu and Joon Son Chung and Arda Senocak},
+  title     = {Seeing Through Touch: Tactile-Driven Visual Localization of Material Regions},
+  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
+  year      = {2026},
+}
+```
+
+## Acknowledgments
+
+This codebase is built upon [TVL (ICML2024)](https://github.com/Max-Fu/tvl).