Chitralekha

Dataset Details

Dataset Version

Some of the fonts do not have proper letters/rendering of different telugu letter combinations. Those have been removed as much as I can find them. If there are any other mistakes that you notice, please raise an issue and I will try my best to look into it

Dataset Description

This extensive dataset, hosted on Huggingface, is a comprehensive resource for Optical Character Recognition (OCR) in the Telugu language, featuring an impressive array of 80+ configurations. Each configuration in this dataset corresponds to a unique font, meticulously curated by Dr. Rakesh Achanta and sourced from his GitHub repository (https://github.com/TeluguOCR/banti_telugu_ocr).

The dataset is specifically designed to support and enhance the development of OCR models, ranging from simple Convolutional Recurrent Neural Network (CRNN) architectures to more advanced systems like trOCR. The versatility of this dataset lies in its large volume and diversity, making it an ideal choice for researchers and developers aiming to build robust OCR systems for the Telugu script.

Key Features:

- Font Diversity: Over 80 unique fonts, each forming a separate configuration, providing a rich variety in text styles and nuances.
- Large Volume: Each configuration contains approximately 800,000 examples, summing up to a vast pool of data for comprehensive training and evaluation.
- Data Split: The dataset is pre-split into training, validation, and test sets, following a 60/20/20 ratio, to facilitate efficient model training and benchmarking.
- Use Cases: Ideal for developing a wide range of OCR models - from basic CRNNs to sophisticated models like trOCR.
- Accessibility: Hosted on Huggingface, ensuring easy access and integration with various machine learning frameworks and tools.

This dataset stands as a testament to Dr. Rakesh Achanta's dedication to enhancing Telugu language processing technologies. It is not just a tool for model development but also a gateway to preserving and digitizing the rich literary heritage of the Telugu language.

Researchers and developers leveraging this dataset are encouraged to adhere to the ethical guidelines of AI research and development, ensuring that the applications developed are for the benefit of language preservation, accessibility, and technological advancement in a responsible manner.

• Fonts Curated by: Dr. Rakesh Achanta
• Shared by: Krishna Sriharsha Gundu
• Data Curated by: Anusha Motamarri
• Language(s) (NLP): Telugu

Ethical Considerations:

Researchers and developers leveraging this dataset are encouraged to adhere to the ethical guidelines of AI research and development. Applications developed using this dataset should prioritize:

• Language preservation and cultural heritage protection
• Improving accessibility of Telugu text for diverse user groups
• Responsible technological advancement in language processing

Dataset Sources [optional]

• Repository: Original Books Dataset