search

microsoft / unilm

Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities
https://aka.ms/GeneralAI
MIT License
19.62k stars 2.5k forks source link

Testing for a custom images #317

Open AarthiAlagammai opened 3 years ago

AarthiAlagammai commented 3 years ago

How to test for a single image with --do-predict without annotations

AarthiAlagammai commented 3 years ago

How to test for a single image with --do-predict without annotations without test.txt

knitemblazor commented 3 years ago

hi, you need to use an ocr engine to extract the text with corresponding bboxes so as to use this flow.

AarthiAlagammai commented 3 years ago

Thanks for the suggestion

On Wed, 12 May 2021 at 10:36, NITISH TOM MICHAEL @.***> wrote:

hi, you need to use an ocr engine to extract the text with corresponding bboxes so as to use this flow.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/microsoft/unilm/issues/317#issuecomment-839438786, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFOPHUBXSX53F77M4SUGI2TTNIEEPANCNFSM422T2JFA .

knitemblazor commented 3 years ago

If you still have doubt refer this repo and especially this file https://github.com/knitemblazor/Multilingual_LayoutLM/blob/main/tesser.py. they wont be releasing the complete flow.

annyWangAn commented 3 years ago

How can I serialize it? Since it is common that there is a line break in the column of the table. How can I keep the text in a colum is not separated?