Despite notable advances in Optical Character Recognition ({OCR}), the accurate recognition of Arabic text remains a persistent challenge due to the script’s cursive morphology, context-dependent letter shaping, and intricate system of diacritics. This study presents a comprehensive comparative analysis of specialized Arabic {OCR} systems and state-of-the-art multimodal vision-language models, including Visual Question Answering ({VQA}) frameworks and large multimodal architectures such as {AIN} and {GPT}-5. By evaluating printed, handwritten, and diacritized Arabic text across diverse public and curated benchmark datasets, we investigate how architectural design, learning objectives, and training paradigms influence recognition fidelity and linguistic robustness. Beyond quantitative assessment, the study examines the broader implications of applying general-purpose multimodal models to language-specific {OCR} tasks, addressing both technological constraints and linguistic sensitivities. Through this investigation, we aim to advance an integrated understanding of how dedicated {OCR} systems and multimodal architectures can converge toward more robust, semantically informed approaches to Arabic document understanding.<\/p>\n","protected":false},"excerpt":{"rendered":"
Despite notable advances in Optical Character Recognition ({OCR}), the accurate recognition of Arabic text remains a persistent challenge due to the script’s cursive morphology, context-dependent letter shaping, and intricate system of diacritics. This study presents a comprehensive comparative analysis of specialized Arabic {OCR} systems and state-of-the-art multimodal vision-language models, including Visual Question Answering ({VQA}) frameworks and large multimodal architectures such as {AIN} and {GPT}-5. By evaluating printed, handwritten, and diacritized […]<\/p>\n","protected":false},"author":12,"featured_media":0,"template":"","meta":{"_acf_changed":false,"footnotes":""},"publication-type":[32],"class_list":["post-35198","publication","type-publication","status-publish","hentry","publication-type-inproceedings"],"acf":[],"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/35198","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication"}],"about":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/types\/publication"}],"author":[{"embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/users\/12"}],"version-history":[{"count":0,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/35198\/revisions"}],"wp:attachment":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/media?parent=35198"}],"wp:term":[{"taxonomy":"publication-type","embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication-type?post=35198"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}