Dragon SDK (Nuance for Voice to Text conversion)

Brief:

Dragon SDK Server Edition (DSS) is a Software Tool which is used to generate an transcription for the speech. It is basically recognize the speech and transcribe the speech in text format. DSS can be integrated with any range of devices like smartphones, tablets, digital recorder, etc. Speech recognition is actually running as a background process so it should not affect any other current process. DSS allows you to create your own custom vocabularies profile for any particular organizations or companies. By using your own personal profile you will get and accurate transcription recognition.

Need:

DSS is mostly used in the voice recognition system. Nowadays you can see in most eCommerce app they provide search an item using the voice. If we take any other example, then it is mostly used for the healthcare system. In healthcare system it will be used as to get the patients information by using their custom healthcare apps. And DSS also provides and special healthcare related speech recognition for the medical vocabularies.

How It Works:

First of all we will see that what is the sytem requirement needed to install the DSS, then we will see how its actually works:

System Requirement


RAM: Minimum 2 GB for 32 bit Windows 7, 8.1 & 10; 4 GB for 64-bit Windows 7, 8.1 & 10 and Windows Server 2008 R2 & 2012
CPU: Intel dual core or equivalent AMD processor. Faster processors yield faster performance.
Free hard disk space: 8 GB if installed from the DVD, 16 GB if you copy the DVD to your hard drive and then install
Supported operating systems: Windows 7, 8.1, 10 (32- and 64-bit); Windows Server 2008 R2 & 201.

Now First you need to create your own custom profile that will actual recognize your voice when any speech come for tarnscription. So create your profile using the DSS tool and save it at your end. It will give you a full profile directory after successfull creation in zip compress format..Now create a new direcoty dragon which can be accessed by your server url. Under that create below structure:

dragon – Create dragon directory
input – Create input subdirectory
output – Create output subdirectory
profile – Create profile subdirectory
training – Create training subdirectory if you did any training on the DSS tool

Now add your custom profile zip file in the profile subdirectory and extract it over there. And if you a training zip then add it in the training directory and extrac it there.

So our next step is to load the speaker or load your profile. So for loading your profile you need to go to the following path and open the loader tool:

C:\Program Files (x86)\Nuance\Dragon SDK Server Edition12\SDK\samples\activex\
vb7

Go to above path and open “openuser.exe” file. Now click on “Browse” button under “Speaker to load” section and select the path for your dragon profile directory. After that click on “Open” button and it will load your speaker and give you success alert “ Speaker Successfully Loaded”. 
Now as next step you need to add a transcription Task in the Transcribe agent tool with input and output directory path. For adding Task into transcribe tool open that tool from following path:

C:\Program Files (x86)\Nuance\Dragon SDK Server Edition12\Program

Open “tagent.exe” file. Now click on Task Menu and and select New. Set the “Input Directory path” to the dragon folder input directory path. Set the “Output Directory path” to the dragon folder output directory path. And Set the “User” path in Location field with the dragon profie directory path. Now click on Ok and the task is assigned to the transcribe agent.

So now we are all done with the above settings. Now if you want to test the transcription just add any speech file(.wav file) in the dragon input directory and it will transcibe the .wav file and create new txt or xml( depends on settings) file in the output directory. 

Podcast

Michael Patterson sat down with the CEO of Boston Byte, Mustapha Shaikh to discuss the significance and rapid digitization of the healthcar...