Search | Contact us
Speech
RDI's Speech Processing Technologies
Arabic NLP
Speech
Summarization
OCR
Arabic LR's
Qur'an Tools
Try Online
Papers

Hafss©, ArabTalk©, Arabic Speech Resources, Arabic Speech Processing, Speech Compression, IVR

Hafss©

Spoken Language/Tajweed Interactive Self Learning Assistant

The goal of this innovative technology is to afford an effective assistance in the self learning of spoken language. Hafss© which has proved effective in helping people learn Tajweed of the glorious Qur’an is also helpful in learning spoken Arabic and other languages as well. In fact, RDI is cooperating with European partners about adapting it to help in learning the less famous spoken European languages like Swedish.

After its first apparition in GITEX 2004, Hafss© proved itself in a large project of building Qur’an web encyclopedia for the Kuwaiti ministry of endowments. The challenging feature about Hafss© is that it can detect and fully diagnose the faults – if any – in the learner’s utterance of given examples. This closes the feedback loop without which the learning process can never be completed.

After GITEX© 2006 and Cairo Int'l Book Fair 2007, an end-user version of Hafss© is released for the Arabic market esp. in the Gulf and Egypt , which is followed by another one with English user-interface, and other versions to emerge in the future. For that version with all its rich material and integrated interactive auxiliary tools, RDI has been awarded the e-Content Award first prize in the e-Learning track from the Egyptian IT Industry Development Agency "ITIDA" for the year 2006.

Moreover, on Sept. 15th, 2007, RDI's Hafss© technology for the interactive learning of spoken language applied on the Tajweed of the glorious Qur'an has been declared first in the 2007 World Summit Award contest for best e-content/e-Learning track where about 850 other candidates from 160 countries had participated.

Where to buy Hafss© ?
(JPG, 468 KB)

Buy Hafss© via the Internet.

Watch AhKam Al-Tajweed episodes taught with the assistance of Hafss(c) by Dr. Ahmed Ragheb

For more detailed info...

Essay on Hafss in PC magazine, Mar. 2007.
(Arabic PDF, 522 KB)

Essay on Hafss in Lughat Al-Assr.
(Arabic WinRAR'ed JPG's, 216 KB)

Arabic Flash movie on Hafss.
(Arabic flash, 2,676 KB)

Brief on Hafss©
(Arabic Doc, 36 KB)

Evaluation Study
(Arabic PDF, 423 KB)

Arabic brochure of Hafss©
(WinZipped images, 1,520 KB)

Hafss© web site; www.rdi-eg.com\Hafss


ArabTalk©

Arabic Text-to-Speech (TTS)

This Arabic TTS from RDI has a hybrid concatenative/parametric speech synthesizer which affords a compact system within few tens of mega bytes of space. Moreover, RDI's automatic large-scale Arabic phonetic transcriptor (diacritizer) is the other distinctive corner stone of ArabTalk©.  

Along with several diacritization options, many speech effects in both male and female voices are available for flexibility and high performance in real-time.  

SDK's for desktop applications, web services, as well as client-server applications are available for developers. Moreover, male/female Arabic TTS speaker data bases can be built for certain real-speakers chosen by the customer.  

For more detailed info..

Arabic Flash movie on Arabic Text-to-Speech.
(Arabic flash, 3,057 KB)

Paper on ArabTalk
(PDF, 96 KB)

Essay on RDI's Arabic Phonetic Transcriptor for Arabic TTS.
(Arabic PDF, 456 KB)


Arabic Speech Processing

Egyptian Telephony ASR DB

Description
Post-processed recordings of around 100 Egyptian speakers from 4 different dialectical regions in Egypt made in Orientel© styles.

Proposed Application(s)
The Up Streaming (via ASR) half of all kinds of Interactive Voice Retrieval (IVR) systems.


Speech Compression

RDI has been known for long of its pioneering CELP vocoding speech compression system that has been awarded the Best-of-GITEX 1996 prize.

In fact, this system that could achieve compression ratios within 15-to-25 at a good quality, has remained for years one of RDI's corner stones for producing its encyclopedic Arabic multimedia that dominated the market in the late 1990's.

After this era, RDI is working on a novel technology for very low bit rate speech coding via a speech re-synthesis process for mobile devices and telecom. applications.


IVR

Introduction

Interactive Voice Response (IVR) is a computerized phone system that enables a person, typically a telephone caller, to make a selection from a voice menu. The selection is made using voice interactions.
The phone system plays RDI’s text to speech enabled voice prompts, and then the person selects the option by spelling it, in turn the person’s voice will be recognized via RDI’s Arabic speech recognition technology.

This interaction allows the individual to communicate with the phone system and thus the computer system either to obtain information or to process a transaction.

IVR can be used in different foundations, such as Call Centers, Banks, Airports, Theaters, etc.

 

Key Features

  • Transforms the phone into a powerful anywhere-anytime information access tool for customers.

  • Reduce call center queues and caller wait times.

  • Extends automated self-service beyond the limits of the traditional Interactive Voice Response.

  • Provides the convenience of touchtone as well as speech-enabled self-service

  • Reduces overall customer service cost by reducing requirements for human agents.

  • Provides high performance call processing and media services.

  • Adopts the open-standards such as VoiceXML, CCXML and Media Resource Control Protocol (MRCP), SIP, and the Web.

  • A software-only solution that runs on off the-shelf hardware and software.

  • Provide a unified development environment that allows businesses to repurpose Web investments to deploy self-service speech applications.

  • The platform can recognize the speech. Also the platform currently includes a TTS voice to convert the text output to a voice message.

     

RDI© - Research and Development International.
Since 1993 - All rights reserved.
Downloads | Jobs
www.rdi-eg.com