CoolFace
Datasetpublic

vrinda2712/SARD

SARD: Synthetic Arabic Recognition Dataset Overview SARD (Synthetic Arabic Recognition Dataset) is a large-scale, synthetically generated dataset designed for training and evaluating Optical Character Recognition (OCR) models for Arabic text. This dataset addresses the critical need for comprehensive Arabic text recognition resources by providing controlled, diverse, and scalable training data that simulates real-world book layouts. Key Features… See the full description on the dataset page: https://huggingface.co/datasets/vrinda2712/SARD.

sourceHugging Faceapache-2.0updated 8mo agoView on Hugging Face
0likes442downloads

vrinda2712/SARD · main · files are served by the source, never re-hosted here