[CVPR 2026] TransPrune: Token Transition Pruning for Efficient Large Vision-Language Model
February 23, 2026 ยท View on GitHub
๐ข News
- [2026-02-23] ๐ Source code for TransPrune are now available!
- [2026-02-21] ๐ Our paper has been accepted by CVPR 2026!
- [2025-07-15] ๐ ArXiv version is released: 2507.20630.
๐ง Install
- Clone this repository
git clone https://github.com/liaolea/TransPrune.git
cd TransPrune
- Install Package
conda create -n transprune python=3.10 -y
conda activate transprune
pip install --upgrade pip # enable PEP 660 support
pip install -e .
๐ Evaluation
We provide scripts to evaluate TransPrune on standard Large Vision-Language Model benchmarks. See .scrpts/transprune/
Cite us
@misc{li2025transprunetokentransitionpruning,
title={TransPrune: Token Transition Pruning for Efficient Large Vision-Language Model},
author={Ao Li and Yuxiang Duan and Jinghui Zhang and Congbo Ma and Yutong Xie and Gustavo Carneiro and Mohammad Yaqub and Hu Wang},
year={2025},
eprint={2507.20630},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2507.20630},
}