600k-ks-ocr: a large-scale synthetic dataset for optical character recognition in kashmiri script

TL;DR

This paper presents the 600K-KS-OCR dataset, a large-scale synthetic resource designed to enable Kashmiri OCR research where data is scarce. It renders approximately 602,000 word-level images at $256\times64$ using three traditional Kashmiri typefaces, augmented with extensive noise, distortion, and background variations to improve robustness. The dataset is distributed in 10 archives (~10.6 GB) and provided in CRNN, TrOCR, CSV, and JSONL formats under CC-BY-4.0, facilitating training, benchmarking, and transfer learning for Kashmiri OCR. By enabling scalable model development and evaluation, it supports digitization and preservation of Kashmiri textual heritage and related low-resource language applications.

Abstract

This technical report presents the 600K-KS-OCR Dataset, a large-scale synthetic corpus comprising approximately 602,000 word-level segmented images designed for training and evaluating optical character recognition systems targeting Kashmiri script. The dataset addresses a critical resource gap for Kashmiri, an endangered Dardic language utilizing a modified Perso-Arabic writing system spoken by approximately seven million people. Each image is rendered at 256x64 pixels with corresponding ground-truth transcriptions provided in multiple formats compatible with CRNN, TrOCR, and generalpurpose machine learning pipelines. The generation methodology incorporates three traditional Kashmiri typefaces, comprehensive data augmentation simulating real-world document degradation, and diverse background textures to enhance model robustness. The dataset is distributed across ten partitioned archives totaling approximately 10.6 GB and is released under the CC-BY-4.0 license to facilitate research in low-resource language optical character recognition.