Add pipeline tag, paper link, and GitHub link to model card

README.md

@@ -1,3 +1,59 @@
----
-license: mit
----
+---
+license: mit
+pipeline_tag: image-feature-extraction
+---
+
+# CoTAP: Semantic Concentration for Self-Supervised Dense Representations Learning
+
+This repository contains the official implementation for the paper "[Semantic Concentration for Self-Supervised Dense Representations Learning](https://huggingface.co/papers/2509.09429)", accepted by IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI 2025).
+
+The code is available at: [https://github.com/KID-7391/CoTAP](https://github.com/KID-7391/CoTAP)
+
+## Overview
+
+CoTAP addresses a key challenge in self-supervised learning: learning dense representations for image patches without the "over-dispersion" phenomenon, where patches from the same instance or category scatter, hindering performance on dense downstream tasks. The method proposes **explicit semantic concentration** through two main components:
+
+1.  **Distilling Patch Correspondences**: To break strict spatial alignment, CoTAP distills patch correspondences, coupled with a novel noise-tolerant ranking loss (an extension of Average Precision loss to continuous targets) to robustly handle noisy and imbalanced pseudo-labels.
+2.  **Object-Aware Filtering**: To discriminate shared patterns from complex scenes, an object-aware filter maps the output space to an object-based space, where patches are represented by learnable prototypes of objects via cross-attention.
+
+This approach effectively learns semantically concentrated dense representations, demonstrating strong empirical results across various tasks.
+
+![CoTAP Pipeline](https://github.com/KID-7391/CoTAP/raw/main/overview-v2.png)
+
+## Checkpoints
+
+| Method | Backbone | Linear Seg (mIoU) | kNN Cls (Acc) | Linear Cls (Acc) | Download |
+| ------ | -------- | ----------------- | ------------- | ---------------- | -------- |
+| DINO+CoTAP    | VIT-S/16 | 50.5 | 78.0 | 79.1 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vits16/dino%2Bcotap.ckpt) |
+| Leopart+CoTAP | VIT-S/16 | 50.4 | 77.3 | 78.6 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vits16/leopart%2Bcotap.ckpt) |
+| iBOT+CoTAP    | VIT-S/16 | 51.6 | 78.4 | 79.4 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vits16/ibot%2Bcotap.ckpt) |
+| Mugs+CoTAP    | VIT-S/16 | 51.6 | 78.9 | 79.4 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vits16/mugs%2Bcotap.ckpt) |
+| DINO+CoTAP    | VIT-B/16 | 52.2 | 79.2 | 80.2 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vitb16/dino%2Bcotap.ckpt) |
+| iBOT+CoTAP    | VIT-B/16 | 53.9 | 80.1 | 80.8 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vitb16/ibot%2Bcotap.ckpt) |
+| Mugs+CoTAP    | VIT-B/16 | 52.1 | 80.8 | 81.7 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vitb16/mugs%2Bcotap.ckpt) |
+| DINO+CoTAP    | VIT-B/8  | 54.7 | 81.6 | 82.1 | [link](https://huggingface.co/KID-7391/CoTAP/blob/main/checkpoints/ours/vitb8/dino%2Bcotap.ckpt) |
+
+## Installation and Usage
+
+For detailed instructions on environment setup, installation, dataset preparation, training, and evaluation, please refer to the [official GitHub repository](https://github.com/KID-7391/CoTAP).
+
+## Citation
+
+If you find this work useful in your research, please cite the following paper:
+
+```bibtex
+@article{wen2025semantic,
+  title={Semantic Concentration for Self-Supervised Dense Representations Learning},
+  author={Wen, Peisong and Xu, Qianqian and Dai, Siran and Cong, Runmin and Huang, Qingming},
+  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)},
+  year={2025},
+  eprint={2509.09429},
+  archivePrefix={arXiv},
+  primaryClass={cs.CV},
+  url={https://arxiv.org/abs/2509.09429},
+}
+```
+
+## Acknowledgements
+
+Our code is based on the official PyTorch implementation of [DINO](https://github.com/facebookresearch/dino), [Leopart](https://github.com/MkuuWaUjinga/leopart), and [STEGO](https://github.com/mhamilton723/STEGO).