1 dataset found

Groups: Word-level Text Detection

Filter Results
  • Total-Text

    Total-Text is a dataset for word-level arbitrary-shaped English text detection, containing 1,255 images for training and 300 images for testing.