MIRACL-VISION Benchmark

What is MIRACL-VISION?

MIRACL-VISION is a large-scale, multilingual visual document retrieval benchmark built by the NVIDIA team, extending the popular MIRACL multilingual text…

Released

2025-05-22

Modality

multimodal

Evaluates

多模态, Multimodal, 语言, Language, VQA, Strong Reasoning, 多模态模型, VLM, 检索能力, Retrieval, 图像理解, Image Understanding, 不支持, Unsupported

Openness

unknown

Importer

OpenCompass Hub

Reported scores

0

Source provenance

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

MIRACL-VISION paper and dataset

• Paper https://arxiv.org/pdf/2505.11651
• Dataset https://huggingface.co/datasets/nvidia/miracl-vision

Which sources cite MIRACL-VISION?

• MIRACL-VISION

Related 多模态, Multimodal, 语言, Language, VQA, Strong Reasoning, 多模态模型, VLM, 检索能力, Retrieval, 图像理解, Image Understanding, 不支持, Unsupported benchmarks

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