Dataset · Hugging Face

TengfeiLiuCoder/MRVBench

Name: MRVBench
Creator: TengfeiLiuCoder
License: cc-by-nc-nd-4.0
Keywords: task_categories:video-text-to-text, task_categories:visual-question-answering, language:en, license:cc-by-nc-nd-4.0, size_categories:n<1K, format:json, modality:image, modality:text, modality:video, library:datasets, library:pandas, library:polars, library:mlcroissant, arxiv:2607.28509, region:us

MRVBench MRVBench is a benchmark for multi-reference image-grounded video captioning. Each example pairs a video with an ordered set of reference images and a caption that locally binds visible subjects, objects, appearances, actions, or scene elements to <Image_n> tags. The reference set may contain distractors that do not correspond to visible video content, enabling joint evaluation of caption factuality, reference selection, phrase-level binding, distractor rejection, and… See the full description on the dataset page: https://huggingface.co/datasets/TengfeiLiuCoder/MRVBench.

View on Hugging FaceReport an issue

Author / organization	TengfeiLiuCoder
License (as declared)	cc-by-nc-nd-4.0
Downloads	383 (as reported by Hugging Face)
Likes	1
Created	31 July 2026
Last updated at source	1 August 2026

video-captioningmulti-imagereference-groundingvisual-question-answeringdistractor-robustness

Embed this entry

<iframe src="https://silytics.si/embed/item/d0f73e2e-9b07-4b2b-a5d2-ddf4d9cb0a0a" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

More from TengfeiLiuCoder

TengfeiLiuCoder/LongAV-CompassDataset

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.