LiMiT (Literal Motion in Text Dataset)

The limit dataset of ~24K sentences that describe literal motion (~14K sentences), and sentences not describing motion or other type of motion (e.g. fictive motion). Senteces were extracted from electronic books categorized as fiction or novels, and a portion from the NetActivity Captions Dataset.

Source: LiMiT

Papers


Paper Code Results Date Stars

Tasks


License


  • Unknown

Modalities


Languages