Fine-tuning LLaMA-3.2-11B on synthetic Manchu images yields an OCR system that reportedly beats a CRNN baseline on real handwritten documents, but test-set-based checkpoint selection and unclear data splits inflate the claim.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
REJECT 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Finetuning Vision-Language Models as OCR Systems for Low-Resource Languages: A Case Study of Manchu
Fine-tuning LLaMA-3.2-11B on synthetic Manchu images yields an OCR system that reportedly beats a CRNN baseline on real handwritten documents, but test-set-based checkpoint selection and unclear data splits inflate the claim.