Touch100k: A large-scale touch-language-vision dataset for touch-centric multimodal representation

Ning Cheng, Jinan Xu, Changhao Guan, Jing Gao, Weihao Wang, You Li, Fandong Meng, Jie Zhou, Bin Fang, Wenjuan Han

Abstract: Highlights•The first 100k tactile dataset with multi-granular descriptions.•A well-designed process for building multimodal datasets.•A method that efficiently captures the links between touch, language, and vision.•Top performance in tactile-related tasks across various benchmarks and settings.