[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [developers] ASR - Arabic Speech Recognition
- To: Development Discussions <developer at arabeyes dot org>
- Subject: Re: [developers] ASR - Arabic Speech Recognition
- From: Ahmad Sayed <ahmad dot ahmadsayed at gmail dot com>
- Date: Mon, 23 Aug 2010 13:47:04 +0200
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:received:in-reply-to :references:date:message-id:subject:from:to:content-type; bh=OyXoSPsrEqJOf5J4HKlthjV0mNdOcr3vcDdFoZHI6n4=; b=ZuD9gwnKlXjCGfWSntbDCo8+kLU+ueGwWfdh4nMNxCuPrBQNhhMw28KxQnQx1hXbu8 1YDhwt2PM4iMeMV14bIyJCiXkFNun6D2/LxtQaUZQNsb3uPnEaKLpymO0GqwvOYTtgPv DwSkb7hwf4vcURh3WuUX2dHAbQOCsPuYLlh28=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=t9znHUD5B5DpdpAAcWGYUYhBZoKjRJ6xXZqaWwwDTx61DKgO5esGgYWItwgUXN2APl R/xZzFrMoSgkGlLcPisR8uoFl480bmwmZnMZhW2sBEhZTRmrikX4CIZY/tfzUj7qc4UQ Zoui1JlTMEOPOz6gFJux7dNOaQCae1XPxE4Vk=
thank you very much for your support i send you a private email to discuss some technical details to avoid disturbing the thread with a long thread.
I think we have a strong foundation to give a push to siragi, anyone available to provide us with the status and update the wiki according to the new updates, unfortunately I am very busy to handle task like this, but i am always ready to provide any technical contributions according to the agreed plan.
On Mon, Aug 23, 2010 at 1:02 PM, Mahmoud Abdel moneim Elgenedy <melgenedy at gmail dot com>
Did you know that you can use HMM in AOCR? i did it in my graduation project using HTK and we got a promising results.
I have present the idea of open source AOCR on developer at arabeyes
from a long time, but the current status of the work is not suitable to release a version as it need more research and some work to convert some matlab files to C.
Kindly I am doing Master Thesis in NLP, I started with speech recognition after doing some achievements the topic halted and switched to OCR, so I would like to share the work i did with the speech recognition with you, it is about 6 months of work and investigation, so I appreciate to contribute it to open source instead of wasting it on my hard drive.
The contribution will be as follow
1 - HTK Models
2 - Small corpus
3 - Trigram language model for the previous corpus.
4 - Some shell scripts and Java code act as utilities.
5 - Some bench-marks and comparison.
6 - small set about 1000 recorded spoken sentence used in the training and testing
Also I am ready to give a quick session with anyone interested to contribute to explain any details the knowledge require is no more than basic programming skills, and shell scripts.
if anyone interested we can set a road map to complete this.
Developer mailing list
Developer at arabeyes dot org