OmniDocBench

What is OmniDocBench?

OmniDocBench is a comprehensive benchmark for evaluating document parsing in real-world scenarios. It includes 981 PDF pages across 9 document types, annotated…

Released

2024-12-02

Modality

multimodal

Evaluates

多模态, Multimodal, 长文本, Long-Context, Document content extraction, 多模态模型, VLM, 长上下文, Long Context, OCR与文档理解, OCR and Document Understanding, 不支持, Unsupported

Openness

unknown

Importer

OpenCompass Hub

Reported scores

0

Source provenance

• Original evidence https://hub.opencompass.org.cn/dataset-detail/OmniDocBench

OmniDocBench paper, code and dataset

• Paper https://arxiv.org/pdf/2412.07626
• Code https://github.com/opendatalab/OmniDocBench
• Dataset https://huggingface.co/datasets/opendatalab/OmniDocBench

Which sources cite OmniDocBench?

• OmniDocBench

Related 多模态, Multimodal, 长文本, Long-Context, Document content extraction, 多模态模型, VLM, 长上下文, Long Context, OCR与文档理解, OCR and Document Understanding, 不支持, Unsupported benchmarks

• 3DCodeBench
• A-Bench
• A-OKVQA
• AbsPyramid
• ActionAtlas
• ActiveFly-Bench
• Ada-LEval
• AECBench