Dataset · Hugging Face

BrightGuo/ProcObject-10K

Name: ProcObject-10K
Creator: BrightGuo
License: cc-by-nc-4.0
Keywords: task_categories:video-text-to-text, task_categories:visual-question-answering, language:en, license:cc-by-nc-4.0, size_categories:10K<n<100K, format:json, modality:tabular, modality:text, library:datasets, library:pandas, library:polars, library:mlcroissant, arxiv:2512.03479, region:us, video-qa

ProcObject-10K Benchmarking Object-Centric Procedural Understanding in Instructional Videos — NeurIPS 2026, Evaluations & Datasets Track · Paper (arXiv:2512.03479) · Code A grounded VideoQA benchmark: a model watches an instructional video clip, answers an open-ended question about how an object's state evolves, and returns the time intervals that support the answer. Answering and grounding are scored separately. 10,522 QA pairs · 1,799 source videos (3,338 clips, 64 h) · 137… See the full description on the dataset page: https://huggingface.co/datasets/BrightGuo/ProcObject-10K.

View on Hugging FaceReport an issue

Author / organization	BrightGuo
License (as declared)	cc-by-nc-4.0
Downloads	348 (as reported by Hugging Face)
Likes	1
Created	5 May 2026
Last updated at source	1 October 2026

video-qatemporal-groundingprocedural-videoobject-centricegocentricmulti-hop-reasoningneedle-in-a-haystack

Embed this entry

<iframe src="https://silytics.si/embed/item/b8432883-9cf2-43cf-860a-95291b7cffca" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Connected in the knowledge graph

mradermacher/ProcObject-Qwen3-VL-4B-GGUFModel · used to train

Explore this entry in the graph

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 Information may be incomplete or out of date.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.