Rhema STS Record
1. Add word in the Name Field
2. Select option for the word text to speech or call phone or both
3. Click add to select a corresponding image for the word as a jpeg file
4. Click the add button to add the word and image to the system
5. click on record to record the relevant voice sound. It need not be a normally recognizable phnetic audio.
6. click the record button will change to stop and when the audio is over, press to stop
7. Continue steps 5 and 6 to add more samples for the same word. 10 times will be a a good sampling given the possibility of difference in sound from the same person.
8. Add any number of words to the system by following steps 1 to 7
9. When the process is completed press the train button. The system will automatically train all the words and audio recorded into the system. NOw you are ready to play.
Rhema STS Play
This is the play back video
1. This is how it will be displayed for the user on the pc monitor or tablet/smart phone screen
2. Click the start button and produce the audio and press stop button
3. The system will retrieve the tagged word and display as shown.
4. By touching the image it will activate the programed sequence, such as dialling a number and passing the related word as voice. If it is text to speech option, then the audio for the text will be played through the speaker.
We have uploaded a short video on youtube at the request of a friend. The link is: http://youtu.be/sugfQgDe9Sg
It is our plan to develop RhemaSTH has a complete unit that would enable a speech and hearing impared person to
1. Train using available speech corpus in their language and phonetic training graphics, as provided by the University of Iowa, on a tablet pc. This would open their vocal chords to express themselves audibly clear to the listner.
2. Provide a blue toothe enabled tactile hardware that would enable them sense the audio, to understand the feedback of what they are saying as well as what others are saying to them. (this is the second stage.)
3. With both integrated into a handheld mobile device, the user can hear phone calls, listen to people and other sounds and be aware of their surroundings.
We have just completed the 1st stage with some training with the students for such handicapped children. I would say that even with little training they were able to pick few words and these guys are totally deaf and dumb. This has surely encouraged us and we are certain many of them will benefit if they could have their own tablet pc with the software loaded. Their energy and tenacity will surely get them to be on par with any normal person.
We propose to provide the whole applications as free open source and hope the final hardware unit will be within US$100 and easily assembled by anyone with some technical know how.
Support and co-operation with people who can share share the wealth of knowledge they have accumulated from past experiences would helpful and welcome..
Sharon Systems has released the US English version of the Rhema Speak to Hear application using the speech corpus provided by CMU Sphinx – Wall Street journal
We have used 10 sample words in this version. Developers can use as many words available in the Corpus and make it available for their clients.
We request you to kindly upload the additions so that it may be available to other users in the US.
Sharon Systems will continue to develop based on avalable speech corpus for other linguistic users. Our ultimate goal is to enable the user to sense the audio feed back of their speech and also sense the audio input from those who communicate with them.
For additional assistance it will be useful to to employ the phonetic graphics made available by the Iowa University. It is useful tool to explain the mouth, tongue, lips and inner chord movements to the learner.
Please download from here at sourceforge.net
Thanks for your support.
Sharon Systems announces the release of the “Rhema STH ver 00.01“ today as an open source software application (FOSS) today.
Rhema STH is designed to help those with speech disability. It helps the individual train their vocal outputs to a Master wave form already recorded and stored in the system. The system is designed to accept readily available speech corpus of any language. Sphinx, the software provided by Carnegie Mellon University, will assimilate the corpus to enable the use of that language in the system.
The first version is being released with the hope and confidence that more talented and interested developers would join in the effort to make a truly viable software application that would help the 416 million people with this disability.
Sharon systems has tested this version with the students of the Dr. MGR school for the speech and hearing impaired, Chennai, India with the kind co-operation of its Principal Mrs Latha Rajendran.
A Tablet PC version for Android, will be released soon. It would help the individual to train with little help from others, to match the wave form and test word recognition by the system.