Greetings! I am working on transcribing pages from scanned pieces, some of which have Latin (more like French, actually) characters with Arabic letters (which were probably added by someone else other than original author to ease the translation). Still, there are situations where there's only Arabic letters and no Latin. I don't know how to read Arabic script, specially if it is handwritten, not typed. I tried running the whole image through the ChatGPT, which is capable of performing OCR and it seemed to help recognize the embedded text. However, I do not have any assurance that what it gives me is correct and not imaginary. What would you recommend to do to accomplish this text recognition task properly? As an example, I want to attach a piece in question, presumably a composition by Bimen Şen Der-Gazaryan...
... for which ChatGPT OCR/Translation gave me the following:
... for which ChatGPT OCR/Translation gave me the following:
- بر خوش نغمه ايـلر او آفت هزار ادر
- ينه عاشق بلبل كیبی دللر هزار ادر
- فدا قيلر دكلئی جانى يار اوغرونده
- آخِرده خار صباح كیبی هر دم هزار ادر
- هم او سڤگلى ديلدار ايله ناز ادر دى
- رضا تشويش ايله دمخانه هزار ادر دى
- جانى وريردى فقيرى حاليم او ديلبر ياره
- فغان كدر كیبی بير عشقى هزار ادر دى
Bir hoş nağme eyler o âfet, hezâr eder
Yine âşık bülbül gibi diller hezâr eder
Feda kılar dükle-i cânı yâr uğrunda
Âhırda hâr sabah gibi her dem hezâr eder
Hem o sevgili dildâr ile naz ederdi
Rızâ, teşvîş ile demhâne hezâr ederdi
Cânı verirdi fakîr-i hâlîm o dilber yâre
Fıġân, keder gibi bir aşkı hezâr ederdi
Would it be possible for someone to validate this translation to the original? Or if there are any other translation suggestions I would appreciate your help. Best regards and thanks!