Enhance model card for ViT-EnsembleAttack

README.md

@@ -1,3 +1,73 @@
 ---
 license: mit
+pipeline_tag: image-classification
+library_name: timm
 ---
+
+# ViT-EnsembleAttack: Augmenting Ensemble Models for Stronger Adversarial Transferability in Vision Transformers
+
+This repository, **TransferAttack**, provides a comprehensive PyTorch framework for various adversarial attacks in image classification. It includes the official implementation of **ViT-EnsembleAttack**, a novel method presented in the paper [ViT-EnsembleAttack: Augmenting Ensemble Models for Stronger Adversarial Transferability in Vision Transformers](https://huggingface.co/papers/2508.12384).
+
+ViT-EnsembleAttack is the first ensemble-based attack method tailored for Vision Transformers (ViTs), designed to significantly enhance adversarial transferability. It achieves this by applying adversarial augmentation to surrogate ViT models using strategies like Multi-head dropping, Attention score scaling, and MLP feature mixing, optimized with Bayesian optimization.
+
+-   📚 [Paper: ViT-EnsembleAttack](https://huggingface.co/papers/2508.12384)
+-   💻 [GitHub Repository](https://github.com/Trustworthy-AI-Group/TransferAttack)
+
+<p align="center">
+  <img src="https://github.com/Trustworthy-AI-Group/TransferAttack/raw/main/figs/overview.png" alt="Overview of TransferAttack Framework" width="700">
+</p>
+
+## ✨ About TransferAttack
+
+TransferAttack is a PyTorch framework developed to boost the adversarial transferability for image classification.
+
+### Why TransferAttack
+
+There are a lot of reasons for TransferAttack, such as:
+
+*   **A benchmark for evaluating new transfer-based attacks**: TransferAttack categorizes existing transfer-based attacks into several types and fairly evaluates various transfer-based attacks under the same setting.
+*   **Evaluate the robustness of deep models**: TransferAttack provides a plug-and-play interface to verify the robustness of models, such as CNNs and ViTs.
+*   **A summary of transfer-based attacks**: TransferAttack reviews numerous transfer-based attacks, making it easy to get the whole picture of transfer-based attacks for practitioners.
+
+## 🚀 Quick Start
+
+### Requirements
+
+Ensure you have the following installed:
+*   Python >= 3.6
+*   PyTorch >= 1.12.1
+*   Torchvision >= 0.13.1
+*   timm >= 0.6.12
+
+You can install the requirements using:
+```bash
+pip install -r requirements.txt
+```
+
+### Usage Example
+
+We adopt an academic-standard ImageNet-compatible dataset comprising 1,000 PNG images for our experiments. Download the data from [GoogleDrive](https://drive.google.com/file/d/1Xx-fJ7_zADhNJRTe7ISB0mdWDT7B8LRr/view?usp=drive_link) or [Huggingface Spaces](https://drive.google.com/file/d/1Xx-fJ7_zADhNJRTe7ISB0mdWDT7B8LRr/view?usp=drive_link) into `/path/to/data`. Then you can execute the attack as follows:
+
+```bash
+python main.py --input_dir ./path/to/data --output_dir adv_data/mifgsm/resnet50 --attack mifgsm --model=resnet50
+python main.py --input_dir ./path/to/data --output_dir adv_data/mifgsm/resnet50 --eval
+```
+
+For detailed information on supported attacks, models, and comprehensive evaluation results (including untargeted and targeted attacks), please refer to the [Attacks and Models section](https://github.com/Trustworthy-AI-Group/TransferAttack#attacks-and-models) and [Evaluation section](https://github.com/Trustworthy-AI-Group/TransferAttack#evaluation) in the GitHub repository.
+
+## 📝 Citation
+
+If you find this work or framework helpful, please consider citing the paper:
+
+```bibtex
+@article{vit-ensembleattack_2024,
+  title={ViT-EnsembleAttack: Augmenting Ensemble Models for Stronger Adversarial Transferability in Vision Transformers},
+  author={[Authors Not Specified in Context]},
+  journal={arXiv preprint arXiv:2508.12384},
+  year={2024}
+}
+```
+
+## 🙏 Contributing to TransferAttack
+
+We welcome suggestions and contributions! For details on how to contribute or to see the full list of contributors and acknowledgements, please refer to the [Contributing section](https://github.com/Trustworthy-AI-Group/TransferAttack#contributing-to-transferattack) in the GitHub repository.