cff-version: 1.2.0
message: "If you use VidTouch, please cite this dataset."
title: "VidTouch: A Multimodal Physical Evidence Benchmark for Material Understanding"
type: dataset
authors:
  - family-names: Wang
    given-names: Yifan
  - family-names: Fu
    given-names: Yu
  - family-names: Wang
    given-names: Jie
  - family-names: Liu
    given-names: Jinduo
  - family-names: Ma
    given-names: Fenglong
year: 2026
repository-code: "https://github.com/cvAQUILA-w/VidTouch_MESA"
url: "https://huggingface.co/datasets/AQUILA-espresso/VidTouch"
license: "CC-BY-4.0"