anke01/Uyghur_OCR_SUST_RUST

--- license: mit task_categories: - optical-character-recognition language: - ug tags: - uyghur - ocr - dataset size_categories: - 1M<n<10M --- # Uyghur OCR Dataset (SUST_RUST) ## Dataset Description This dataset contains Uyghur language OCR (Optical Character Recognition) training data. ### Dataset Summary - **Language:** Uyghur (维吾尔语) - **Task:** Optical Character Recognition (OCR) - **Size:** ~12GB (compressed) ### Supported Tasks - Optical Character Recognition (OCR) - Text Recognition - Multi-lingual OCR Research ### Languages The dataset primarily contains Uyghur language text images. ## Dataset Structure The dataset is provided as a compressed zip file (`ug_dataset.zip`) containing: - Training images - Corresponding labels/annotations ## Usage ```python from datasets import load_dataset dataset = load_dataset("anke01/Uyghur_OCR_SUST_RUST") ``` ## Dataset Creation ### Source Data This dataset was collected and prepared for Uyghur OCR research. ## Additional Information ### Dataset Curators SUST_RUST Team ### Licensing Information MIT License ### Citation If you use this dataset, please cite: ```bibtex @dataset{uyghur_ocr_sust_rust, title={Uyghur OCR Dataset}, author={SUST_RUST Team}, year={2024}, publisher={Hugging Face}, url={https://huggingface.co/datasets/anke01/Uyghur_OCR_SUST_RUST} } ```