search

microsoft / unilm

Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities
https://aka.ms/GeneralAI
MIT License
19.49k stars 2.48k forks source link

TROCR for Rectangular images #1282

Open abhiagwl4262 opened 1 year ago

abhiagwl4262 commented 1 year ago

Is TrOCR good choice for handwritten text Recognition for images with large W/H ratio of 5-6. E.g. 600*100 size images ?

TrOCR resizes input image to square image of 384*384 which distorts the image a lot.

o-evgeny commented 11 months ago

same question

abhiagwl4262 commented 11 months ago

I think its there - https://github.com/microsoft/unilm/blob/b60c741f746877293bb85eed6806736fc8fa0ffd/trocr/data_aug.py#L34 But I have not got the time to explore it any further.