MMCSBench: A Fine-Grained Benchmark for Large Vision-Language Models in Camouflage Scenes

(Updating slowly, but keep going)

Dataset Path

Hugging Face: https://huggingface.co/datasets/zjswsz/MMCSBench

Kaggle: https://www.kaggle.com/datasets/f2218284b51011e4e27d4c4d8b41eff771d0ec734c535a7a0b94bb2a02058a46?resource=download

How to use

TODO

Results

Note: This repository contains the official implementation of the MMCSBench benchmark presented at NeurIPS 2025. For the latest updates and announcements, please watch this repository and check our project page.