💫 Vulcan · Class-Specific ViT Derivation 💫

🌀 What is Vulcan?

🚀 Vulcan is a novel approach for deriving compact, class-specific Vision Transformers (ViTs) tailored for resource-constrained edge devices. 🎯 Given a pre-trained base ViT, Vulcan can derive a lightweight ViTs that focus on recognizing the target classes.

📂 Project Structure

Folder/File	Description
src/data	Stores sub-task definitions and intermediate experimental results.
src/dataset	Dataset loading, processing, and augmentation utilities.
src/engine	Core training and evaluation pipelines.
src/method	Core implementations of Vulcan, including CCNC and TNNR losses, adaptive configuration, and structured pruning.
src/model	Model definitions and loading utilities for ViT and Swin backbones.
src/scripts	Shell scripts for running Vulcan experiments with different models, tasks, and pruning configurations.
src/utils	General utility functions for profiling, FLOPs/parameter calculation, memory analysis, and training support.
src/main.py	Main entry point to run Vulcan, including post-training and pruning.

🚀 Quick Start

1. Clone the repository

First, clone the NuWa project to your local machine:

git clone https://github.com/xxx/vulcan.git
cd vulcan/scripts

2. Install required dependencies
3. Run the pipeline

./vulcan_base.sh

📎 Citation

If you find this code useful, please cite our paper:

@article{wei2026vulcan,
  title={Vulcan: Crafting compact class-specific Vision Transformers for edge intelligence},
  author={Wei, Ziteng and He, Qiang and Chen, Feifei and Duan, Ranjie and Li, Xiaodan and Li, Bin and Chen, Yuefeng and Xue, Hui and Jin, Hai and Yang, Yun},
  journal={International Conference on Learning Representations},
  year={2026}
}