@inproceedings{zong-qiu-2024-gaokao,
    title = "{GAOKAO}-{MM}: A {C}hinese Human-Level Benchmark for Multimodal Models Evaluation",
    author = "Zong, Yi  and
      Qiu, Xipeng",
    editor = "Ku, Lun-Wei  and
      Martins, Andre  and
      Srikumar, Vivek",
    booktitle = "Findings of the Association for Computational Linguistics: ACL 2024",
    month = aug,
    year = "2024",
    address = "Bangkok, Thailand",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/add-emnlp-2024-awards/2024.findings-acl.521/",
    doi = "10.18653/v1/2024.findings-acl.521",
    pages = "8817--8825",
    abstract = "The Large Vision-Language Models (LVLMs) have demonstrated great abilities in image perception and language understanding. However, existing datasets either focus solely on primary perception abilities and commonsense knowledge, or have a low level of text comprehension difficulty, which are insufficient to reflect the comprehensive capabilities of LVLMs, particularly in terms of Chinese language proficiency. We propose GAOKAO-MM, a multimodal benchmark based on the Chinese College Entrance Examination (GAOKAO), comprising of 8 subjects and 12 types of images, such as diagrams, function graphs, maps and photos. GAOKAO-MM derives from native Chinese context and sets human-level requirements for the model`s abilities, including perception, understanding, knowledge and reasoning. We evaluate 10 LVLMs and find that the accuracies of all of them are lower than 50{\%}, with GPT-4-Vision (48.1{\%}), Qwen-VL-Plus (41.2{\%}) and Gemini-Pro-Vision (35.1{\%}) ranking in the top three positions. The results of our multi-dimension analysis indicate that LVLMs have moderate distance towards Artificial General Intelligence (AGI) and provide insights facilitating the development of multilingual LVLMs. The dataset and evaluation code are available through: https://github.com/OpenMOSS/GAOKAO-MM"
}