search

cs-chan / Total-Text-Dataset

Total Text Dataset. It consists of 1555 images with more than 3 different text orientations: Horizontal, Multi-Oriented, and Curved, one of a kind.
BSD 3-Clause "New" or "Revised" License
747 stars 140 forks source link

The result of end-to-end recognition. #23

Closed MhLiao closed 5 years ago

MhLiao commented 5 years ago

@cs-chan It seems that there is a mistake in the result of end-to-end recognition. In the Neurocomputing2019 paper, the authors say that "Total-Text [33] contains 744 horizontal text images for testing", which means that they did not evaluate the curved samples. Moreover, it is only a text recognition method which does not involve detection, instead of an end-to-end recognition method which involves both detection and recognition.

cs-chan commented 5 years ago

Thank you for your information. We have checked and confirmed it. In this case, we have removed the paper in the table to avoid confusion.