🔨

[1_2___2.1.1] MMOCR visualization 도구의 사용방법을 확인한다.

상태

Done

담당

마감일

2022/11/24

요약

Detection 이미지, recognition 이미지, Augmentation 파이프라인을 통과한 이미지, 통과하지 않은 이미지 모두를 확인하면 성공

선행 태스크

선행 태스크 상태

후행 태스크

관련 마일스톤

2 more properties

•

분석해 본 결과

◦

train_dataloader 만 정의되어 있으면 된다.

▪

Based on this tool, users can easily verify if the annotation of a custom dataset is correct. Also, you can verify if the data augmentation strategies are running as you expected by modifying train_pipeline in the configuration file.

▪

train_dataloader 은 train_pipeline 을 사용하는데, train_pipeline 에는 runtime augmentation 등이 정의되어 있다. 

•

조금 더 자세히 이해하기 위해서는 아래 두 문서를 참고하면 좋다.

Useful Tools - MMOCR 1.0.0rc3 documentation

MMOCR provides a dataset visualization tool tools/analysis_tools/browse_datasets.py to help users troubleshoot possible dataset-related problems. You just need to specify the path to the training config and the tool will automatically plots the images transformed by corresponding data pipelines with the GT labels.

https://mmocr.readthedocs.io/en/dev-1.x/user_guides/useful_tools.html#dataset-visualization-tool

Data Transforms and Pipeline - MMOCR 1.0.0rc3 documentation

In the design of MMOCR, dataset construction and preparation are decoupled. That is, dataset construction classes such as OCRDataset are responsible for loading and parsing annotation files; while data transforms further apply data preprocessing, augmentation, formatting, and other related functions. Currently, there are five types of data transforms implemented in MMOCR, as shown in the following table.

https://mmocr.readthedocs.io/en/dev-1.x/basic_concepts/transforms.html