SlideShare ist ein Scribd-Unternehmen logo
1 von 115
AOCR Arabic Optical Character Recognition ABDEL RAHMAN GHAREEB KASEM ADEL SALAH ABU SEREEA MAHMOUD ABDEL MONEIM ABDEL MONEIM MAHMOUD MOHAMMED ABDEL WAHAB
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Introduction ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Preprocessing ,[object Object],[object Object],[object Object],[object Object],[object Object]
Preprocessing Problem of tilted image 1. Image rotation
Preprocessing 1. Process rotated image
[object Object],Preprocessing 1. Process rotated image
[object Object],Preprocessing 1. Process rotated image
[object Object],Preprocessing 1. Process rotated image
[object Object],Preprocessing 1. Process rotated image
[object Object],Preprocessing 1. Process rotated image
Preprocessing 1. Process rotated image ,[object Object],Clear zeros Clear zeros Mean value 0.2*Mean value
Preprocessing 1. Process rotated image GRAY Scale Vs. Black/White  in Rotation process Original image Gray scale Black/White
Preprocessing ,[object Object],[object Object],[object Object],[object Object],[object Object]
Preprocessing ,[object Object],[object Object],[object Object],[object Object]
[object Object],Preprocessing ,[object Object]
[object Object],Preprocessing ,[object Object]
[object Object],Preprocessing ,[object Object]
[object Object],Preprocessing
Preprocessing ,[object Object],[object Object],[object Object],[object Object],[object Object]
Preprocessing 3.  Image enhancement
3.  Image enhancement  Preprocessing ,[object Object],By morphology operations
[object Object],Apply  Image Enhancement  operations on  small  images not  large  image بسم الله الرحمن الرحيم الله أكبر الله أكبر الله أكبر لا إله الا الله والله أكبر Large Image X ,[object Object],بسم الله الرحمن الرحيم الله أكبر الله أكبر الله أكبر لا إله الا الله والله أكبر
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Feature   Extraction   الله   اكبر
[object Object],[object Object],[object Object],[object Object],[object Object],we select features such that:
Satisfaction of the previous points ,[object Object],n1 n3 n4 n6 n5 n2 n7 Feature vector محمد رسول الله
[object Object],[object Object]
[object Object],الصلاة overlap
(1)  Background Count   ,[object Object],background Foreground النجاح
[object Object],Example: d1 d3 d2 d3  d2  d1  Feature vector of the selected slide Two pixels with on overlap
Feature Figure
(2) Baseline Count   ,[object Object]
Example: Baseline   No. of black pixels above baseline (X1) No. of black pixels below baseline (X2) Two pixels with on overlap   Thinning X2  X1  Feature vector
Feature Figure
(3) Centroid   ,[object Object],Example: Cx  Cy  Feature vector Two pixels with on overlap
(4) Cross Count ,[object Object],Example: 2   Feature vector Two pixels with on overlap
(5) Euclidean distance ,[object Object]
Baseline   Euclidean distance above baseline D1 Euclidean distance below baseline D2 Example: Thinning One pixel without overlap   D2  D1  Feature vector
Feature Figure
(6) Horizontal histogram   ,[object Object],Calculate Histogram Example: Four pixels with one overlap
Feature Figure
(7) Vertical histogram   ,[object Object],Example: X2   X1  Feature vector Two pixels with one overlap
Feature Figure
( 8) Weighted vertical histogram ,[object Object]
Example: weight vector 1 -1 X2   X1  Feature vector Two pixels with one overlap
Feature Figure
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Implementation of AOCR Based HMM Using HTK ,[object Object],[object Object],[object Object]
Data preparation ,[object Object],[object Object],[object Object],[object Object],[object Object]
The Task Grammar ,[object Object],[object Object]
Isolated AOCR Grammar ,[object Object],[object Object],[object Object],[object Object]
Connected AOCR Grammar ,[object Object],[object Object],[object Object],[object Object]
Why  Grammar? Start a1 a2 a124 a125 a3 End
How is it created? ,[object Object],Grammar Word Net (  Wdnet ) HParse
The Dictionary ,[object Object],???
The Dictionary
Recording the Data Feature extraction Transformer (Image)  2-D signal 1-D vector .wav
Creating the Transcription Files ,[object Object],[object Object]
Word level MLF #! MLF! # "*/1.lab" فصل . "*/2.lab" في الفرق بين الخالق والمخلوق . "*/3.lab" وما ابراهيم وآل ابراهيم الحنفاء والأنبياء فهم . "*/4.lab" يعلمون انه لا بد من الفرق بين الخالق والمخلوق . . . فصل في الفرق بين الخالق والمخلوق وما ابراهيم وآل ابراهيم الحنفاء والأنبياء فهم يعلمون انه لا بد من الفرق بين الخالق والمخلوق
Phone  level MLF #! MLF! # "*/1.lab" a74 a51 a88 . "*/2.lab" a74 a108 a123 a1 a86 a75 a38 a77 a123 #! MLF! # "*/1.lab" فصل . "*/2.lab" في الفرق بين الخالق والمخلوق . "*/3.lab" وما ابراهيم وآل ابراهيم الحنفاء والأنبياء فهم . "*/4.lab" يعلمون انه لا بد من الفرق بين الخالق والمخلوق . .
Coding the Data HCOPY MFCC Files S0001.mfc S0002.mfc S0003.mfc etc.. Wave form files ٍٍ S0001.wav S0002.wav S0003.wav etc.. Configuration File Script File
Creating Monophone HMMs ,[object Object],[object Object]
Creating Monophone HMMs ,[object Object],[object Object],[object Object]
The Prototype ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Initialization Process Proto  Vfloors Proto HCompV hmm0
Initialized prototype ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Vfloors Contents  ,[object Object],[object Object],[object Object]
Creating initialized Models a125 a2 a1 Initialized  model hmmdefs ~o <VecSize> 39 <MFCC_0_D_A> Initialized  proto
Creating Macros File Vfloors file ~o <VecSize> 39 <MFCC_0_D_A> Vfloors file
Re-estimation Process   Hmmdefs macros HERest Initialized  Proto HCompV Hmmdefs macros Training Files MFc Files Phones level Transcription monophones
Recognition Process  Hvite Trained  Models Test Files Word Network wnet The dictioary dict Reconized words
Recognizer Evaluation HResults Reference Transcription Reconized Transcription Accuracy
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Experimental Results
1- Main Problem ,[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],Discrete Input images. (failed) Continuous Input a continuous wave form (Succeeded) DATA Input to HTK
2- Isolated Character Recognition ,[object Object],[object Object],[object Object]
2-1  Single Size (16)- Single Font (Simplified Arabic Fixed) ,[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],116 100% 35 No. of  Models 99.14 % Accuracy
[object Object],[object Object],11 99.14% 3 No. of  States 96.55 % Accuracy
[object Object],[object Object]
2-2  Multi-Sizes Character Recognition ,[object Object],[object Object]
2-3  Variable Lengths Character Recognition   ,[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],[object Object],[object Object],Accuracy=100%
3-  Connected Character Recognition   ,[object Object],[object Object],[object Object],[object Object]
3-1  Single Size (16)- Single Font (Simplified Arabic Fixed) ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object]
[object Object]
3-2  Parameter Optimization ,[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Word Level 85.36% Line Level Level 84.99%   Accuracy
[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object]
[object Object],[object Object],[object Object],11 95.02% 8 Number of states 92.52% Accuracy =
[object Object],[object Object],[object Object],[object Object],0  1  0  0  0 0  0.7  0.3  0  0 0 0  0.6  0.4  0 ------------------------------and so on.
[object Object],[object Object],[object Object],0.5 93.92% 0.4 0.6 Overlapping Ratio = 91.70% 92.52% Accuracy =
[object Object],Vertical histogram 96.97% Max. Accuracy Feature Type 95.96% 2-D histogram 87.16% Euclidean distance 91.51% Cross count 95.75% Weighted histogram 89.70% Baseline count 91.61% Background count
3-3  Multi-Sizes Character Recognition ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],16 96.97% 18 14 Font size 76.21% 79.74% Accuracy
[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Accuracy = 92.15%
3-4 F eature concatenation ,[object Object],No scale 5 84.09% 4 4 Scale vertical histogram)= 4.2 5.57 Window size = 69.02% 77.17% Accuracy =
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Future works   ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
[object Object],[object Object],[object Object]
Main contents ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]
Automatic Form Recognition ,[object Object],بنــك مصــر شيك رقم :  .......................... اسم المصرف اليه : ................. المبلغ بالارقام : ................  المبلغ بالحروف : .................. امضاء ...................
[object Object],[object Object],بسم الله
[object Object],[object Object],بسم الله
 

Weitere ähnliche Inhalte

Was ist angesagt?

An optimal and progressive algorithm for skyline queries slide
An optimal and progressive algorithm for skyline queries slideAn optimal and progressive algorithm for skyline queries slide
An optimal and progressive algorithm for skyline queries slideWooSung Choi
 
Design of frequency selective surface comprising of dipoles using artificial ...
Design of frequency selective surface comprising of dipoles using artificial ...Design of frequency selective surface comprising of dipoles using artificial ...
Design of frequency selective surface comprising of dipoles using artificial ...IJAAS Team
 
Computer graphics
Computer graphics Computer graphics
Computer graphics shafiq sangi
 
How to tune a query - ODTUG 2012
How to tune a query - ODTUG 2012How to tune a query - ODTUG 2012
How to tune a query - ODTUG 2012Connor McDonald
 
Introduction to Gura Programming Language
Introduction to Gura Programming LanguageIntroduction to Gura Programming Language
Introduction to Gura Programming LanguageYutaka Saito
 
Electrical Engineering Exam Help
Electrical Engineering Exam HelpElectrical Engineering Exam Help
Electrical Engineering Exam HelpLive Exam Helper
 
Kaggle Winning Solution Xgboost algorithm -- Let us learn from its author
Kaggle Winning Solution Xgboost algorithm -- Let us learn from its authorKaggle Winning Solution Xgboost algorithm -- Let us learn from its author
Kaggle Winning Solution Xgboost algorithm -- Let us learn from its authorVivian S. Zhang
 
Optimization
OptimizationOptimization
OptimizationManas Das
 
Introduction of Xgboost
Introduction of XgboostIntroduction of Xgboost
Introduction of Xgboostmichiaki ito
 

Was ist angesagt? (18)

Skyline queries
Skyline queriesSkyline queries
Skyline queries
 
An optimal and progressive algorithm for skyline queries slide
An optimal and progressive algorithm for skyline queries slideAn optimal and progressive algorithm for skyline queries slide
An optimal and progressive algorithm for skyline queries slide
 
8th Semester, Computer Science and Information Science (2012-June/) Question ...
8th Semester, Computer Science and Information Science (2012-June/) Question ...8th Semester, Computer Science and Information Science (2012-June/) Question ...
8th Semester, Computer Science and Information Science (2012-June/) Question ...
 
Cgm Lab Manual
Cgm Lab ManualCgm Lab Manual
Cgm Lab Manual
 
Design of frequency selective surface comprising of dipoles using artificial ...
Design of frequency selective surface comprising of dipoles using artificial ...Design of frequency selective surface comprising of dipoles using artificial ...
Design of frequency selective surface comprising of dipoles using artificial ...
 
Computer graphics
Computer graphics Computer graphics
Computer graphics
 
Cg lab cse-v (1) (1)
Cg lab cse-v (1) (1)Cg lab cse-v (1) (1)
Cg lab cse-v (1) (1)
 
How to tune a query - ODTUG 2012
How to tune a query - ODTUG 2012How to tune a query - ODTUG 2012
How to tune a query - ODTUG 2012
 
Oct.22nd.Presentation.Final
Oct.22nd.Presentation.FinalOct.22nd.Presentation.Final
Oct.22nd.Presentation.Final
 
Introduction to Gura Programming Language
Introduction to Gura Programming LanguageIntroduction to Gura Programming Language
Introduction to Gura Programming Language
 
8th Semester Computer Science (2013-June) Question Papers
8th Semester Computer Science (2013-June) Question Papers8th Semester Computer Science (2013-June) Question Papers
8th Semester Computer Science (2013-June) Question Papers
 
Electrical Engineering Exam Help
Electrical Engineering Exam HelpElectrical Engineering Exam Help
Electrical Engineering Exam Help
 
ARPS Architecture
ARPS ArchitectureARPS Architecture
ARPS Architecture
 
Kaggle Winning Solution Xgboost algorithm -- Let us learn from its author
Kaggle Winning Solution Xgboost algorithm -- Let us learn from its authorKaggle Winning Solution Xgboost algorithm -- Let us learn from its author
Kaggle Winning Solution Xgboost algorithm -- Let us learn from its author
 
Optimization
OptimizationOptimization
Optimization
 
Introduction of Xgboost
Introduction of XgboostIntroduction of Xgboost
Introduction of Xgboost
 
Cpp tutorial
Cpp tutorialCpp tutorial
Cpp tutorial
 
E04612529
E04612529E04612529
E04612529
 

Ähnlich wie Aocr Hmm Presentation

Polymer Brush Data Processor
Polymer Brush Data ProcessorPolymer Brush Data Processor
Polymer Brush Data ProcessorCory Bethrant
 
Optimization in Programming languages
Optimization in Programming languagesOptimization in Programming languages
Optimization in Programming languagesAnkit Pandey
 
Wave File Features Extraction using Reduced LBP
Wave File Features Extraction using Reduced LBP Wave File Features Extraction using Reduced LBP
Wave File Features Extraction using Reduced LBP IJECEIAES
 
닷넷 개발자를 위한 패턴이야기
닷넷 개발자를 위한 패턴이야기닷넷 개발자를 위한 패턴이야기
닷넷 개발자를 위한 패턴이야기YoungSu Son
 
IRJET - Automatic Licence Plate Detection and Recognition
IRJET -  	  Automatic Licence Plate Detection and RecognitionIRJET -  	  Automatic Licence Plate Detection and Recognition
IRJET - Automatic Licence Plate Detection and RecognitionIRJET Journal
 
C++ Notes PPT.ppt
C++ Notes PPT.pptC++ Notes PPT.ppt
C++ Notes PPT.pptAlpha474815
 
lecture1.ppt
lecture1.pptlecture1.ppt
lecture1.pptSagarDR5
 
Text Detection and Recognition in Natural Images
Text Detection and Recognition in Natural ImagesText Detection and Recognition in Natural Images
Text Detection and Recognition in Natural ImagesIRJET Journal
 
Beholding the giant pyramid of application development; why Ajax applications...
Beholding the giant pyramid of application development; why Ajax applications...Beholding the giant pyramid of application development; why Ajax applications...
Beholding the giant pyramid of application development; why Ajax applications...Javeline B.V.
 
B Eng Final Year Project Presentation
B Eng Final Year Project PresentationB Eng Final Year Project Presentation
B Eng Final Year Project Presentationjesujoseph
 
2007 Tidc India Profiling
2007 Tidc India Profiling2007 Tidc India Profiling
2007 Tidc India Profilingdanrinkes
 
EELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docx
EELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docxEELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docx
EELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docxtoltonkendal
 
Making Custom Oscilloscope Measurements
Making Custom Oscilloscope MeasurementsMaking Custom Oscilloscope Measurements
Making Custom Oscilloscope Measurementsteledynelecroy
 
Build Your Own 3D Scanner: 3D Scanning with Structured Lighting
Build Your Own 3D Scanner: 3D Scanning with Structured LightingBuild Your Own 3D Scanner: 3D Scanning with Structured Lighting
Build Your Own 3D Scanner: 3D Scanning with Structured LightingDouglas Lanman
 
Project 2: Baseband Data Communication
Project 2: Baseband Data CommunicationProject 2: Baseband Data Communication
Project 2: Baseband Data CommunicationDanish Bangash
 
Lossless Data Compression Using Rice Algorithm Based On Curve Fitting Technique
Lossless Data Compression Using Rice Algorithm Based On Curve Fitting TechniqueLossless Data Compression Using Rice Algorithm Based On Curve Fitting Technique
Lossless Data Compression Using Rice Algorithm Based On Curve Fitting TechniqueIRJET Journal
 
VCE Unit 01 (2).pptx
VCE Unit 01 (2).pptxVCE Unit 01 (2).pptx
VCE Unit 01 (2).pptxskilljiolms
 
(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...
(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...
(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...Naoki Shibata
 

Ähnlich wie Aocr Hmm Presentation (20)

Polymer Brush Data Processor
Polymer Brush Data ProcessorPolymer Brush Data Processor
Polymer Brush Data Processor
 
Optimization in Programming languages
Optimization in Programming languagesOptimization in Programming languages
Optimization in Programming languages
 
Dsp lab manual 15 11-2016
Dsp lab manual 15 11-2016Dsp lab manual 15 11-2016
Dsp lab manual 15 11-2016
 
ECG BASED REPORT.
ECG BASED REPORT.ECG BASED REPORT.
ECG BASED REPORT.
 
Wave File Features Extraction using Reduced LBP
Wave File Features Extraction using Reduced LBP Wave File Features Extraction using Reduced LBP
Wave File Features Extraction using Reduced LBP
 
닷넷 개발자를 위한 패턴이야기
닷넷 개발자를 위한 패턴이야기닷넷 개발자를 위한 패턴이야기
닷넷 개발자를 위한 패턴이야기
 
IRJET - Automatic Licence Plate Detection and Recognition
IRJET -  	  Automatic Licence Plate Detection and RecognitionIRJET -  	  Automatic Licence Plate Detection and Recognition
IRJET - Automatic Licence Plate Detection and Recognition
 
C++ Notes PPT.ppt
C++ Notes PPT.pptC++ Notes PPT.ppt
C++ Notes PPT.ppt
 
lecture1.ppt
lecture1.pptlecture1.ppt
lecture1.ppt
 
Text Detection and Recognition in Natural Images
Text Detection and Recognition in Natural ImagesText Detection and Recognition in Natural Images
Text Detection and Recognition in Natural Images
 
Beholding the giant pyramid of application development; why Ajax applications...
Beholding the giant pyramid of application development; why Ajax applications...Beholding the giant pyramid of application development; why Ajax applications...
Beholding the giant pyramid of application development; why Ajax applications...
 
B Eng Final Year Project Presentation
B Eng Final Year Project PresentationB Eng Final Year Project Presentation
B Eng Final Year Project Presentation
 
2007 Tidc India Profiling
2007 Tidc India Profiling2007 Tidc India Profiling
2007 Tidc India Profiling
 
EELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docx
EELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docxEELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docx
EELE 5331 Digital ASIC DesignLab ManualDr. Yushi Zhou.docx
 
Making Custom Oscilloscope Measurements
Making Custom Oscilloscope MeasurementsMaking Custom Oscilloscope Measurements
Making Custom Oscilloscope Measurements
 
Build Your Own 3D Scanner: 3D Scanning with Structured Lighting
Build Your Own 3D Scanner: 3D Scanning with Structured LightingBuild Your Own 3D Scanner: 3D Scanning with Structured Lighting
Build Your Own 3D Scanner: 3D Scanning with Structured Lighting
 
Project 2: Baseband Data Communication
Project 2: Baseband Data CommunicationProject 2: Baseband Data Communication
Project 2: Baseband Data Communication
 
Lossless Data Compression Using Rice Algorithm Based On Curve Fitting Technique
Lossless Data Compression Using Rice Algorithm Based On Curve Fitting TechniqueLossless Data Compression Using Rice Algorithm Based On Curve Fitting Technique
Lossless Data Compression Using Rice Algorithm Based On Curve Fitting Technique
 
VCE Unit 01 (2).pptx
VCE Unit 01 (2).pptxVCE Unit 01 (2).pptx
VCE Unit 01 (2).pptx
 
(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...
(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...
(Slides) Efficient Evaluation Methods of Elementary Functions Suitable for SI...
 

Aocr Hmm Presentation

  • 1. AOCR Arabic Optical Character Recognition ABDEL RAHMAN GHAREEB KASEM ADEL SALAH ABU SEREEA MAHMOUD ABDEL MONEIM ABDEL MONEIM MAHMOUD MOHAMMED ABDEL WAHAB
  • 2.
  • 3.
  • 4.
  • 5.
  • 6.
  • 7. Preprocessing Problem of tilted image 1. Image rotation
  • 8. Preprocessing 1. Process rotated image
  • 9.
  • 10.
  • 11.
  • 12.
  • 13.
  • 14.
  • 15. Preprocessing 1. Process rotated image GRAY Scale Vs. Black/White in Rotation process Original image Gray scale Black/White
  • 16.
  • 17.
  • 18.
  • 19.
  • 20.
  • 21.
  • 22.
  • 23. Preprocessing 3. Image enhancement
  • 24.
  • 25.
  • 26.
  • 27. Feature Extraction الله اكبر
  • 28.
  • 29.
  • 30.
  • 31.
  • 32.
  • 33.
  • 35.
  • 36. Example: Baseline No. of black pixels above baseline (X1) No. of black pixels below baseline (X2) Two pixels with on overlap Thinning X2 X1 Feature vector
  • 38.
  • 39.
  • 40.
  • 41. Baseline Euclidean distance above baseline D1 Euclidean distance below baseline D2 Example: Thinning One pixel without overlap D2 D1 Feature vector
  • 43.
  • 45.
  • 47.
  • 48. Example: weight vector 1 -1 X2 X1 Feature vector Two pixels with one overlap
  • 50.
  • 51.
  • 52.
  • 53.
  • 54.
  • 55.
  • 56. Why Grammar? Start a1 a2 a124 a125 a3 End
  • 57.
  • 58.
  • 60. Recording the Data Feature extraction Transformer (Image) 2-D signal 1-D vector .wav
  • 61.
  • 62. Word level MLF #! MLF! # &quot;*/1.lab&quot; فصل . &quot;*/2.lab&quot; في الفرق بين الخالق والمخلوق . &quot;*/3.lab&quot; وما ابراهيم وآل ابراهيم الحنفاء والأنبياء فهم . &quot;*/4.lab&quot; يعلمون انه لا بد من الفرق بين الخالق والمخلوق . . . فصل في الفرق بين الخالق والمخلوق وما ابراهيم وآل ابراهيم الحنفاء والأنبياء فهم يعلمون انه لا بد من الفرق بين الخالق والمخلوق
  • 63. Phone level MLF #! MLF! # &quot;*/1.lab&quot; a74 a51 a88 . &quot;*/2.lab&quot; a74 a108 a123 a1 a86 a75 a38 a77 a123 #! MLF! # &quot;*/1.lab&quot; فصل . &quot;*/2.lab&quot; في الفرق بين الخالق والمخلوق . &quot;*/3.lab&quot; وما ابراهيم وآل ابراهيم الحنفاء والأنبياء فهم . &quot;*/4.lab&quot; يعلمون انه لا بد من الفرق بين الخالق والمخلوق . .
  • 64. Coding the Data HCOPY MFCC Files S0001.mfc S0002.mfc S0003.mfc etc.. Wave form files ٍٍ S0001.wav S0002.wav S0003.wav etc.. Configuration File Script File
  • 65.
  • 66.
  • 67.
  • 68. Initialization Process Proto Vfloors Proto HCompV hmm0
  • 69.
  • 70.
  • 71. Creating initialized Models a125 a2 a1 Initialized model hmmdefs ~o <VecSize> 39 <MFCC_0_D_A> Initialized proto
  • 72. Creating Macros File Vfloors file ~o <VecSize> 39 <MFCC_0_D_A> Vfloors file
  • 73. Re-estimation Process Hmmdefs macros HERest Initialized Proto HCompV Hmmdefs macros Training Files MFc Files Phones level Transcription monophones
  • 74. Recognition Process Hvite Trained Models Test Files Word Network wnet The dictioary dict Reconized words
  • 75. Recognizer Evaluation HResults Reference Transcription Reconized Transcription Accuracy
  • 76.
  • 78.
  • 79.
  • 80.
  • 81.
  • 82.
  • 83.
  • 84.
  • 85.
  • 86.
  • 87.
  • 88.
  • 89.
  • 90.
  • 91.
  • 92.
  • 93.
  • 94.
  • 95.
  • 96.
  • 97.
  • 98.
  • 99.
  • 100.
  • 101.
  • 102.
  • 103.
  • 104.
  • 105.
  • 106.
  • 107.
  • 108.
  • 109.
  • 110.
  • 111.
  • 112.
  • 113.
  • 114.
  • 115.