• No results found

Real-Time Distributed Speech Transcription System

N/A
N/A
Protected

Academic year: 2021

Share "Real-Time Distributed Speech Transcription System"

Copied!
8
0
0

Loading.... (view fulltext now)

Full text

(1)

Real-Time Distributed Speech Transcription System

NeSToR

(2)

Application

“Nestor” is a cutting-edge transcription (real-time and suspended) solution. The audio-signal distribution technology allows speedy, secure and reliable transcription of speeches, meetings, lectures, conferences and court proceedings. “Nestor” is easy to install and operate. The flexibility build-in into “Nestor” allows easy customization, based on the needs of a customer.

Benefits

The system has a unique functionality and the following distinctive features:

Speed - It takes only 5 minutes from the end of a speech/event to the edited and spell-checked transcription.

Cross-check - Ability to link the final text with the corresponding audio fragment ensures maximum correctness of transcription

Reliability - The system is absolutely safe in case of network disconnections and power shut-downs

easy to use – convenient Word environment

Brief description of the complex

Audio signal coming into the system is recorded into the hard disc. Then it is divided into fragments and distributed between any given number typists for actual transcription. The text fragments are automatically assembled in the single audio-text document which can be saved and printed out.

Case study

“Nestor” is already installed and successfully working in:

Administration of the President of Russia State Duma (Russian Parliament)

Government of Russia

Constitutional Court of Russia Supreme Court of Belorus Local authorities In the pipe-line Parliament of Mauritius            

(3)

Work Flow

Transcriptions in 3 steps

1 - Audio signal transferred to Nestor

2 - Audio divided into segments and distributed among typist/work groups 3 - Assembled audio-linked file, ready to print

(4)

Nestor consists of a number of:

Specialized working stations (SWS)

SWS «Server»

SWS «Recording station»

Automated working stations (AWS)

AWS «Recording manager» AWS «Administrator»

AWS «Typist» (editor and operator) AWS «Software import module»

Software module in charge sound file into the server

Functionality

Digital recording through 8 mono channels or 4 stereo channels from one sound card

Acoustic and visual quality control

Automatic transfer of fragments between operators and editors. The length of the fragments is adjustable to the typist’s comfort

Automatic assembly of the fragments into a single document

Audio-text file as a final result, where clicking on a particular text fragment brings the corresponding audio

Based on MS Word editor Automatic spell-check

Playback advanced options (loop, fastening/slowing without audio distortion) Foot pedal for playback control

Text messages exchange between typists and message history

Different fragment dispatching strategies (manual capturing, automatic dispatching)

Audio and text archiving on the server Wealth of project statistics

Comments to each project “name”, “place”

Final document revision with simultaneous playback of the corresponding audio fragments

Hardware and software recording station and server back-up

                      

(5)

Server

Server – is a key element of the system. All components of the “Nestor” areconnected through the Server. Software designed for the Server includes DBMS Microsoft SQL express. Client’s working stations makes request to DBMS of the Server by means of oLeDB. All significant data which are important for the work of the Complex is securely saved on the Server and the back up Server.

Sound files

Information about the recordings Information about the projects

Segmentation of the recordings on fragments Final reports of speeches

Sound fragments

Back-up module

Back-up module is installed on the server and synchronizes all data with it. In case the server falls out, the system continuous working using back-up stations.

Record station

Nestor’s recording station includes the following functional components:

Input audio card Drivers for audio card

Subsystem of sound data reading Subsystem of back-up data storage Subsystem of data transferring

Nestor’s recording station combines the functions of digital recorder and a network application. Recording station has the following functionality:

Audio recording on the hard disc

Transfer of any additional data from an operator to the server Network connection with the server

Back-up of the received data

Synchronous transferring of the data in the server Mono/stereo switch                 

(6)

Record manager

Record manager is the interface for recording station

Remote control over the recording station

Ability to support several recording stations at a time Spectrum analyzer for any channel

Channels grouping for multi-channel sessions

Control over the record transfer from the recording station to the server Sound card adjustment and acoustic control on any channel

Administrator

Administrator is the team leader, directing the operation of the team of typists. Administrator’s working station allows to:

Define roles of the working stations in the complex Delete records and projects

Create projects and groups of operators Make settings for the project

Switch operators and editors between projects

View the statistics of the operators’ and editors’ work for a day

Control the degree of completeness of the final document and fragments Send informational messages to any of the components of the complex View the information on the complex

Typist

AWS “operator” features a user friendly interface and is designed as a module for MS Word text editor.

editor

editor’s role is to assemble the final document from the text fragments prepared by individual typists. editor’s working stating can be used as a Typist’s station as well.

Typist

Typist transcribes audio fragments into a transcript. Typists can work on several projects at the same time, switching between main and back-up servers, exchange messages with other Typists, editors and Administrators.

Foot pedal, automatic saving, hot keys and many other functions are available.

              

(7)

Possible configurations

4 users (3 AWS “operator”, 1 AWS “editor” combined with AWS “operator”, AWS “Administrator”)

6 users (5 AWS “operator”, 1 AWS “editor” combined with AWS “operator”, AWS “Administrator”)

9 users ( 7 AWS “operator”, 2 AWS “editor” combined with AWS “operator”, AWS “Administrator”)

11 ( 9 AWS “operator”, 2 AWS “editor” combined with AWS “operator”, AWS “Administrator”) *

* - System can be upgraded with more AWS upon request

   

(8)

Russia, 4 Krasutskogo str, Saint Petersburg

tel. +7 812 331-0665, fax: +7 812 327-9297

References

Related documents

In order to study halocin production among halophilic archaea and bacteria in different types of hypersaline environments, 82 archaeal and bacterial strains obtained

Their system lies on an archi- tecture consisting of five layers: (i) the data source layer represented by the available Twitter APIs, (ii) the ETL layer (Extract, Transform, and

Of this total, 80 are male characters (51 %) and 78 instances are female characters The principal characters, characters and 58 female characters. This means that the most

• "Why can't we have a relational database of toxic chemicals and occupational diseases to store and query information similar to ones used by companies to manage data

In humans, BTX injections are used in masticatory muscles (masseter, temporalis, lateral pterygoid and rarely medial pterygoid) for several indications such as

(2009): “Sustainable Net Present Value: A value-based approach to sustainable investment appraisal”, Annual Meeting of the Academy of Management, Chicago, 7 – 11 August 2009..

However normal USB devices like wireless keyboard mouse plugs, bluetooth wireless internet dongles will continue to work.. without

The fact that countries with high or increasing yield growth rates (class 1 – 4) and with stagnating or decreasing yields (respectively class 5 and 6) can be found in all NPP