Thai OCR Evaluation Images and Text from Open-Source Websites

A dataset for evaluating Optical Character Recognition models on Thai language text. It contains images and corresponding textual data derived from various open-source websites. The dataset was created by openthaigpt and last updated in September 2024.

Use Cases

• Evaluate OCR model accuracy on Thai text extracted from images sourced from open websites.
• Benchmark multilingual OCR systems using the Thai language images and their corresponding ground truth text.
• Assess OCR performance on handwritten Thai text, as indicated by the 'Handwritten' tag.

Strengths

• Includes both image and text modalities, providing ground truth for OCR evaluation.
• Tagged as multilingual and specifically for the Thai language, targeting a specific linguistic domain.
• Last updated in September 2024, indicating recent maintenance.

Limitations

• Unknown sample size, making it difficult to assess statistical significance for model evaluation.
• Data fields and column structure are not specified, limiting understanding of the data format.
• Source is described as 'various open-source websites', which may introduce inconsistency in data quality and collection methods.

Provenance

Source

openthaigpt on Hugging Face.

Collection Method

Images and textual data derived from various open-source websites.

Freshness

Last updated on 2024-09-30.

Geography

Region tag indicates 'us', but specific geographic coverage of source material is unknown.