Skip to content

Latest commit

 

History

History
26 lines (20 loc) · 1.02 KB

README.md

File metadata and controls

26 lines (20 loc) · 1.02 KB

Data format example

This implementation requires the input data in the following format:

Training/Validation data

  1. txt_file txt file: multi-line with format: ImageFileName,TextLabel. e.g. data/example_gt.txt
  2. img_root folder: image files corresponding to ImageFileName. e.g. data/example_imgs

Testing data

  1. img_folder folder: image files.

Or use lmdb format dataset and modified the corresponding config file.

Lmdb data

Base on lmdb dataset in deep text recognition benchmark: Source repository

Download lmdb dataset: Here

Unzip and modify folder below and use config file config_lmdb.json to train: (Can custom sub folder in folder training and val by fix select_data in config file)

dataset/data_lmdb_release/
    training/
        MJ_train
        ST
    val/
        MJ_test
        MJ_valid