Google Web Speech API Implementation

Report
Google Web Speech API
Implementation
Case Study: English Skill Online Practice
Prajaks Jitngernmadan
Faculty of Informatics, Burapha University
Agenda
 Objectives
 Introduction
 Technologies
 System Design
 System Implementation
 Conclusion
 References
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
2
Objectives
 To design an automatic free-text
answer correcting system architecture
 To apply Google Web Speech API for
speech recognition
 To design and develop a simple online
English training system prototype
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
3
Introduction
 Language is one of the key successful
factors.
 Multi-cultural society.
 Common Language  English
 AEC in 2015
 Most of Thais lack of English speaking
and listening skill.
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
4
Introduction
 Thai Characters
 Shy
 Modest
  especially when it comes to
English
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
5
Introduction
 The online English training system has
to be
 Supporting the users to practice
translating and speaking skill with
real-time feedback.
 Easy to access  web-based.
 Encouraging the users interactively.
 Providing necessary information.
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
6
Technologies
 How to correct the answer
 Text comparison method
 Similar Text Algorithm
Input Answer
:: user
similar_text()
% similarity
Given key
:: admin
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
7
Technologies
 How to recognize the speech
 Web Speech API
 <input lang=”en” x-webkit-speech>
 Google Chrome version 25+
Google Web Speech API
Input Speech
:: user
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
Output Text
8
System Design
 Translating Practice Structure
Compared by
Similar Text
Solve-1
Input
answer
user
Solve-2
select the
best result
Highest Score
Calculate
Solve-3
Threshold
>= 80%?
no
wrong
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
yes
correct
9
System Design
 Speaking Practice Structure
speak
Web browser
Chrome
Web Speech API
(over Internet)
Google
Server
user
Text from voice
Convert to text
Translating Process
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
10
System Design
 Overall System Structure
user
Internet
Google Chrome 25+
Application
Server
Database
admin
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
11
System Design
 Database Design
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
12
System Implementation
 First Prototype Approach
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
13
System Implementation
 First Prototype Approach
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
14
Conclusion
 Hands-on English system (English Skill
Online Practice)
 Web-based application
 Designed and developed under the
conceptual interaction and didactics
 System gives the responses within a
minute
 for speaking, the response time depends on the
Internet connection speed
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
15
Conclusion
 Future Works
 Improve the answer correcting method
 Redesign the user interface for more interactive
and user-friendly
 Test and collect data of learning improvement for
statistical analysis
 Acknowledgment
 Faculty of Informatics at Burapha University
 Computer Graphics and Multimedia Research
Laboratory
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
16
References
 Similar text, (2013). PHP Manuals, http://php.net/manual/en/function.similartext.php. Last Accessed: December 15, 2013.
 Web Speech API, (2013). https://dvcs.w3. org/hg/speech-api/raw-file/tip/
speechapi.html. Last Accessed: December 15, 2013
 jQuery, (2013). http://jquery.com/. Last Accessed: December 15, 2013.
 ภาษาอังกฤษ, (2013). http://th.wikipedia.org/ wiki/ภาษาอังกฤษ Last Accessed: Dec 15, 2013
 EF EPI Country Rankings, (2013) . http://www.ef.co.th/epi/ Last Accessed Dec
15, 2013
 Google Web Speech API, (2013).
http://commondatastorage.googleapis.com/io-2013/presentations/
4057%20Web%20Speech%20API%20creates%20Interactive%20Experiences
%20-%20Google%20I-O%202013%20Codelab.pdf, Last Accessed Dec 15,
2013
 Michael F. McTear, (2002). Spoken dialogue technology: enabling the
conversational user interface, ACM Computing Surveys (CSUR) Volume 34
Issue 1, pp 90-169.
 Araki, M. and Takegoshi, D. (2012). Computer Software and Applications
Conference (COMP SAC), pp 195 - 202.
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
17
Thank you for your
attention!
Google Web Speech API Implementation :: Prajaks Jitngernmadan :: BUU
18

similar documents