VRTool User Guide

VR Tool is a transcription tool implemented using IBM Watson Speech to text Conversion.

File Upload Flow Diagram 

Watson Account Configuring in VRTool

VRTool uses IBM Watson Speech to Text service to transcribe the dictations.To start using VRTool user has to create a IBM Watson account and configure it in the VRTool settings screen.

 Profile Creation

VRTool -> Profile Manager -> Create Profile

For Each dictation provider create create a profile in VR tool to start using the tool. Once the profile is created user can start sending files to IBM Watson to get it transcribed. But this will only give you the out of the box accuracy of IBM Watson engine. It is recommended to train the profile by providing enough audio and transcripts of the dictation provider.

 Training

Once the profile is created , it can be trained using archive audio and transcripts form the same dictation provider. For Better text quality, IBM Watson demands two types of training – Language and Acoustic. This can be achieved from this Tool (Please see the screenshots below).

Language Training

During Language training Watson familiarize with the way user dictates the file . It also collects any new words that is not there in the IBM Watson dictionary. To do language training you need to collect enough number of transcripts of the same dictation provider.

 Acoustic Training

During acoustic training IBM Watson profile learns about the acoustic characteristic of the speaker.While creating profile user has select the sampling rate of audio. During training and live use provide same type file.

Sending Audio for Transcription

Once training is completed , VR tool can be used to send files to IBM Watson for transcription. It is possible to queue multiple files for transcription. After uploading files user can check the status of files in FileManager section.

File Manager

File Manager shows complete list of files done using the tool. User can open the VR Transcribed file in the editor. Transcribed texts are color coded with accuracy.After complete the editing finalize the file to remove color coding.

Editing Tool

From file manager screen user can open the transcripts in the editing tool. Editing tool embeds MS Word in it along with the audio player. User can navigate through the file and hear the audio corresponds to text and vice versa. Color coding helps user to concentrate on the places needs attention.