LLM-Driven AutoML for Cross-Lingual Handwritten OCR: Closed-Loop Neural Architecture Search with GPT-5, GPT-4o, and Claude Sonnet 4
Computer Vision and Pattern RecognitionArtificial IntelligenceMachine Learning
Summary
The authors built a system where AI models like GPT-5, GPT-4o, and Claude Sonnet 4 design handwriting recognition neural networks by themselves. These AI models try different designs, train them, check how well they work, and improve them over time without any human help. They tested this approach on Arabic, Persian, and English handwriting and got high accuracy, over 93%, sometimes reaching 98.1%. The models are also fast and do not need special tweaks or preparation of the data. This shows that large language models can be useful for automatically creating good handwriting recognition systems in multiple languages.
Authors
Mobina Kashaniyan, Amirhossein Ghassemi, Nasser Mozayani
Abstract
We present a fully automated closed-loop AutoML framework that uses GPT-5, GPT-4o, and Claude Sonnet 4 as autonomous neural architecture designers for cross-lingual handwritten optical character recognition. Each large language model independently generates, trains, evaluates, and iteratively refines neural network architectures using performance feedback from previous trials. The framework is evaluated on Arabic, Persian, and English handwriting datasets through 270 independent experiments. It consistently discovers accurate and computationally efficient models without manual architecture design, domain-specific preprocessing, or hyperparameter tuning. The generated models achieve mean test accuracies above 93 percent, a best accuracy of 98.1 percent, and inference latency between 41 and 44 milliseconds. The results demonstrate that large language models can function as effective AutoML agents for neural architecture search, enabling scalable, script-adaptive, and reproducible handwriting recognition across languages.