Skip to content

Page-wise text recognition with lower-supervision line data models

License

Notifications You must be signed in to change notification settings

UB-Mannheim/party

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

party

party is **PA**ge-wise **R**ecognition of **T**ext-y. It is a replacement for conventional text recognizers in ATR system using the baseline+bounding polygon line data model where it eliminates the need for bounding polygons.

Installation

$ pip install .

Training

Options are largely identical to those offered by ketos train, apart from supporting only XML training data:

$ party -d cuda train -f xml --workers 32 -t train.lst -e val.lst

About

Page-wise text recognition with lower-supervision line data models

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%