MTabVQA

该数据集名为MTabVQA，是一个专门设计用来评估多模态模型在处理以图像形式呈现的多个表格中进行多跳推理能力的基准测试。它包含了3,745个复杂的问题-答案对，这些问题需要对视觉渲染的表格图像进行推理。MTabVQA还包括了从几个涉及多表连接的复杂查询的文本到SQL数据集衍生出的子数据集，并且它使用渲染过程将表格数据转换为图像，以供视觉推理任务使用。其任务是进行多表格视觉问题解答。

The dataset named MTabVQA is a benchmark specifically designed to evaluate the multi-hop reasoning ability of multimodal models when handling multiple tables presented in image format. It contains 3,745 complex question-answer pairs that require reasoning over visually rendered table images. MTabVQA also includes sub-datasets derived from several text-to-SQL datasets involving complex queries with multi-table joins, and it employs a rendering process to convert tabular data into images for visual reasoning tasks. The core task of this benchmark is multi-table visual question answering.