@KartikC14 For line segmentation, please see ihdia.iiit.ac.in/LineTR/
English
Ravi Kiran S
2.7K posts
Ravi Kiran S
@vikataravi
Associate Prof, IIIT-H, India | Vision@BharatGen | iHub-Data | Alum: UW, IISc | Enjoy interdisciplinary problems involving multimedia | Meme fan | 🇮🇳
Hyderabad, India Katılım Mart 2013
511 Takip Edilen1.6K Takipçiler
@vikataravi Very cool!!
Does the text-line segmentation stage of the system require manuscript specific finetuning? Would love to collaborate with you regarding this
English
A breakthrough for Indic manuscript OCR.
Introducing UniLipi - a unified Sanskrit manuscript OCR system capable of decoding across 13 scripts.
English
@vikataravi @iiit_hyderabad is there a website where we can upload grantha manuscripts and get unicode grantha/ devanagari output with this?
English
@vikataravi Very Good!!
Congratulations!!
Is it available on internet?
English
@sanely_Prashant it's mostly about accessibility. if manuscripts from these lipi can be sourced, they can also be supported.
English
@RajeevBAgarwal it's mostly about accessibility. if manuscripts from these regions can be sourced, they can also be supported.
English
@vikataravi Congratulations!
By the way
- why no script for Gujarat, Rajasthan, Madhya Pradesh regions?
- May like to edit Bengali to Bangla
English
@_gottilla_ it's mostly about accessibility. if manuscripts from these lipi can be sourced, they can also be supported.
English
@vikataravi Can it decode 12th century TuLu lipi? There are thousands of these manuscripts in the Udupi, Karnataka.
English
@DAkkhar it's mostly about accessibility. if manuscripts from these lipi can be sourced, they can also be supported.
English
@vikataravi Takri Lipi for Jammu & Himachal is missing. Takri has two main variants Jammu and Chamba
Eesti
@sanjeevku it's mostly about accessibility. if manuscripts from these lipi can be sourced, they can also be supported.
English
@vikataravi This is excellent. It would be good if a similar effort could be done for Brahmi and Prakrit scripts also. 👏🏼👏🏼
English
@tshrinivas Namaste. UniLipi = Unified approach for multiple lipi.
Filipino
@vikataravi Introduction of 'UniLipi' is a breakthrough indeed for Indic manuscript OCR. Calling it 'Ekalipi' (एकालिपि) might make it more 'Indic.'
English
@G_U_C_C_I__COOL yes. output is akshara-based. it can be rendered in any language which has a matching font.
English
@vikataravi Can in do the other way round ? Devanagari to kannada ?
English
@centerofright it's mostly about accessibility. if manuscripts from these regions can be sourced, they can also be included.
English
@vikataravi Have a query - that entire belt between Guj, Raj, MP and Chattisgarh - nothing found - Surprising
English
@yajnadevam @meghkalyan Not yet. But can extend to Brahmi. Will require some seed annotated data.
English
QAM