# TVL (Touch-Vision-Language) dataset

From the catalogue. Listed by one researcher from one source page.

- id: `tvl-touch-vision-language-dataset`
- kind: language and reasoning annotation
- robot or device: handheld 3D printed device with DIGIT sensor and Logitech BRIO webcam (class: sensor only)
- how collected: handheld in the wild plus robot-collected SSVTP; GPT-4V pseudo-labels
- size as stated: 43,741 image-touch pairs (SSVTP 4,587; HCT 39,154)
- hours: not stated | hours only claimed: none | episodes: not stated
- year: 2024
- organisation: UC Berkeley, Meta AI, TU Dresden, CeTI
- licence: not stated (class: not stated)
- access: open (class: open)
- link (dataset): https://tactile-vlm.github.io/
- note: On Hugging Face as mlfu7/Touch-Vision-Language-Dataset.

Confirm the licence at the link before relying on it.