목차

• Language
• Vision
  • Image
• Vision + Language
  • Video
  • Captioning
    • Text Captioning

이 글에서는 Deep learning을 사용하는 다양한 분야의 데이터셋을 최대한 정리한다.

Language

Vision

Image

Vision + Language

Video

Howto100M

YFCC100M

YouCook2

Something-Something V2 (SSv2)(174 classes, 168.9K/24.7K video)

Captioning

Text Captioning

TextVQA

Towards VQA Models That Can Read

ST-VQA

Scene Text Visual Question Answering

TextCaps

TextCaps: a Dataset for Image Captioning with Reading Comprehension