Posts Tagged ‘Urdu’

Dynamic Subtitle Translator

Friday, August 20th, 2010

If you frequently watch movies with the sub-titles, but have problem finding the sub-titles in your native language, then you can easily translate the existing/available sub-titles file to language of your choice.

Dynamic Subtitle Translator is a web-application which translate the Subtitle from one language to other language. On the back end, it uses the powerful Google Translation service to do the actual translation. Once the automatic translation is done, you can also manually edit/refine the translation and then export the translated subtitle back to the original format.

Though the translated text is not 100% accurate (as it is only machine translation), but at least 70% of the translated text is correct, plus if you have courage and patience, you can manually edit the translated text and make a perfect translation and share it with others.

One of the nice, and my main inspiration behind this tool, is the sub-title translation to the Urdu. It’s not perfect, but using the language transliteration feature, you can easily tweak sub-titles in minutes. Just imagine watching foreign movie with Urdu sub-titles.

Go, give it a try and share you feedback and comments. The service link is:
http://www.syedgakbar.com/products/dst/

Hindi to English and Hindi to Urdu Transliteration

Saturday, February 6th, 2010

Urdu and Hindi are two of the most popular languages of sub-continent. The speaker of these two languages combined are more than a billion. However the one major difference between these two languages is how these are written. The Hindi language is written using Devanagari script, while Urdu language is written in Urdu script (oriented from Persian and Arabic).

As a native speaker and reader of Urdu, the Hindi written in Devanagari script would appear to me as alien language even though I would recognize 70 to 80 percent of it when some one actually read that for me. One day tired of this complexity, I started exploring the automatic transliteration options. As always, some very good work is already done on this and here are some of the interesting sites I found for transliteration:
http://www.crulp.org/software/langproc/h2utransliterator.html
http://www.puran.info/HUMT/index.html
http://translate.malerkotla.co.in/

Though these are great, but there are some problems with these transliterations services like:
1. As there is not direct mapping of some characters between these two languages so most of the time transliteration of such characters is not correct e.g. devanagari character ? (za) in Urdu could be mapped to to ?(zay), ?(zoad), ?(zal) and ?(zoin) characters depending on context
2. Some of these transliteration services, just provide a limited text box for transliteration and may be too limiting depending on what you want to convert. Plus you also have to visit their site and copy and paste your content to get the transliteration done.
3. Third and most important, these are slow. I think this is due to the reason that they are actually doing word look-up in the database/dictionary for the mapping. This brings excellent results, but response is slow.

To overcome some of these problems, I have written my own Hindi to English transliteration engine. This transliteration engine follows mixture of ITRANS and IAST Hindi transliteration schemes (to produce more human readable text).

For the Hindi to Urdu transliteration, I decided to use Google Indic Transliteration API. So for this mapping, I first convert the Hindi to Roman and then pass this to Google Transliteration API for Roman to Urdu transliteration. The final results are amazing. You get almost ready to use transliteration. The other plus points are that it’s very fast (that’s why I’m fan of Google), and you can do this transliteration on any page using the Dynamic Language Tools bookmarklet. You can read more about this here:
http://www.syedgakbar.com/products/web/

Here is a quick demo of this service in action:
Hindi to English Transliteration