Home

Awesome

DetToolChain<img src="images/unicorn.png" width="3%">: Unleashing Detection Ability of MLLMs

[Paper]

<p align="center" width="100%"> <a ><img src="images/main.png" alt="overview" style="width: 100%; min-width: 300px; display: block; margin: auto;"></a> </p>

Visual Processing Prompts

<p align="center" width="60%"> <a ><img src="images/vis.png" alt="overview" style="width: 70%; min-width: 300px; display: block; margin: auto;"></a> </p>

Getting Started

1. Installation

Git clone our repository:

git clone https://github.com/yixuan730/DetToolChain
cd DetToolChain

2. Run Cantor Demo

python main.py 

Citation

@article{wu2024dettoolchain,
  title={Dettoolchain: A new prompting paradigm to unleash detection ability of MLLM},
  author={Wu, Yixuan and Wang, Yizhou and Tang, Shixiang and Wu, Wenhao and He, Tong and Ouyang, Wanli and Wu, Jian and Torr, Philip},
  journal={arXiv preprint arXiv:2403.12488},
  year={2024}
}