search

lamhoangtung / LineHTR

Line-level Handwritten Text Recognition (HTR) system implemented with TensorFlow.
https://towardsdatascience.com/2326a3487cd5
MIT License
74 stars 23 forks source link

Evaluation #14

Open AICampB4 opened 3 years ago

AICampB4 commented 3 years ago

Hello. Thank for your efficient work. But I've read your readme and create a labels.json which contains 'img_path': 'context' and I used the IAM offline lines dataset for the input data. But I don't know why the result is very bad: image This is my systems informations:

  1. Versions
    • TensorFlow version: 1.x
    • Python version: 3.7.12
    • Operating system: Google colab I'm a newbie and in research of this domain, so could you please help me which part I did wrong. Thank you in advance.
AICampB4 commented 3 years ago

Sorry, I don't how why the image do not upload correctly. So I copied the result: Character error rate: 19.624679%. Text line accuracy: 4.307692%. Word error rate: 369.538462% Character error rate not improved No more improvement since 32 epochs. Training stopped.