Improve dataset card: Add task category, paper link, and enhance description

README.md

@@ -1,7 +1,44 @@
 ---
 license: cc-by-nc-4.0
+task_categories:
+  - image-segmentation
+tags:
+  - text-image-forgery
+  - data-synthesis
 ---
 
-Official implementation of Fourier Series-based Tampering Synthesis (FSTS) from our NeurIPS 2025 paper “Toward Real-world Text Image Forgery Localization: Structured and Interpretable Data Synthesis”
+# FSTS Dataset
 
-Project Page: https://github.com/ZeqinYu/FSTS
+Official implementation of Fourier Series-based Tampering Synthesis (FSTS) from our NeurIPS 2025 paper “[Toward Real-world Text Image Forgery Localization: Structured and Interpretable Data Synthesis](https://huggingface.co/papers/2511.12658)”.
+
+Code/GitHub Repository: https://github.com/ZeqinYu/FSTS
+
+## Abstract
+
+Existing Text Image Forgery Localization (T-IFL) methods often suffer from poor generalization due to the limited scale of real-world datasets and the distribution gap caused by synthetic data that fails to capture the complexity of real-world tampering. To tackle this issue, we propose Fourier Series-based Tampering Synthesis (FSTS), a structured and interpretable framework for synthesizing tampered text images. FSTS first collects 16,750 real-world tampering instances from five representative tampering types, using a structured pipeline that records human-performed editing traces via multi-format logs (e.g., video, PSD, and editing logs). By analyzing these collected parameters and identifying recurring behavioral patterns at both individual and population levels, we formulate a hierarchical modeling framework. Specifically, each individual tampering parameter is represented as a compact combination of basis operation-parameter configurations, while the population-level distribution is constructed by aggregating these behaviors. Since this formulation draws inspiration from the Fourier series, it enables an interpretable approximation using basis functions and their learned weights. By sampling from this modeled distribution, FSTS synthesizes diverse and realistic training data that better reflect real-world forgery traces. Extensive experiments across four evaluation protocols demonstrate that models trained with FSTS data achieve significantly improved generalization on real-world datasets.
+
+## News
+*   **[2025.11.18]** 🚀 We have released the [arXiv version](https://arxiv.org/abs/2511.12658), with additional experiments and analyses included in the appendix.
+*   **[2025.11.14]** 🚀 The FSTS dataset is publicly available on the following platforms: [Hugging Face](https://huggingface.co/datasets/Zegkim/FSTS).
+*   **[2025.11.07]** 🔥 We have released video samples that showcase the implementation pipeline of FSTS.
+
+## Examples of FSTS Synthesis Process
+Below are several examples (slow version) showing the *automatic synthesis pipeline* of FSTS, which collects and models real-world tampering parameters to synthesize samples with the _replacement_ operation.
+
+### Example1
+https://github.com/user-attachments/assets/d627d363-8983-44f4-affe-0b87f9ea7343
+
+### Example2
+https://github.com/user-attachments/assets/062d52df-9823-4b7d-a685-48b1f53ae90b
+
+### Example3
+https://github.com/user-attachments/assets/d31570d1-694a-4a76-af18-a51f59d45d85
+
+## Citation
+```bibtex
+@inproceedings{yutoward,
+  title={Toward Real-world Text Image Forgery Localization: Structured and Interpretable Data Synthesis},
+  author={Yu, Zeqin and Xie, Haotao and Zhang, Jian and Ni, Jiangqun and Su, Wenkang and Huang, Jiwu},
+  booktitle={The Thirty-ninth Annual Conference on Neural Information Processing Systems Datasets and Benchmarks Track}
+}
+```