search

ying09 / TextFuseNet

A PyTorch implementation of "TextFuseNet: Scene Text Detection with Richer Fused Features".
MIT License
476 stars 123 forks source link

datasets format #52

Open oym050922021 opened 3 years ago

oym050922021 commented 3 years ago

hi, what is the format of training-datasets ground truth ? Is it similar to the ground truth of detection or semantic segmentation? Should the location of each word be labeled? Can this model be used to do semantic segmentation tasks only? Thank you very much!

Real-YeJ commented 3 years ago

@oym050922021

  1. details about the format pls refer to COCO dataset.
  2. we only use the semantic segmentation branch to extract global context