MVL-SIB Benchmark

What is MVL-SIB?

MVL-SIB is a multilingual dataset that provides image-sentence pairs spanning 205 languages and 7 topical categories (entertainment, geography, health…

Released

2025-02-18

Modality

multimodal

Evaluates

多模态, Multimodal, 理解, Understanding, 语言, Language, 多模态模型, VLM, 图像理解, Image Understanding, 语言理解, Comprehension, 不支持, Unsupported

Openness

unknown

Importer

OpenCompass Hub

Reported scores

0

Source provenance

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

MVL-SIB paper and dataset

• Paper https://arxiv.org/abs/2502.12852
• Dataset https://huggingface.co/datasets/Davlan/sib200
• Dataset https://huggingface.co/datasets/WueNLP/mvl-sib
• Dataset https://huggingface.co/datasets/WueNLP/mvl-sib200

Which sources cite MVL-SIB?

• MVL-SIB

Related 多模态, Multimodal, 理解, Understanding, 语言, Language, 多模态模型, VLM, 图像理解, Image Understanding, 语言理解, Comprehension, 不支持, Unsupported benchmarks

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