Ankush Goyal Karan Singh Prof. K. Gopalan
During the course Of this presentation…
 
OBJECTIVE T o design a  real-time  speaker verification system
SPEAKER RECOGNITION Speaker Identification Who is it? 1:N Is he really Ankush? Speaker Verification 1:1 ?? I am Ankush!
PHASES OF SYSTEM Enrollment Phase Authentication Phase
Feature Extraction Threshold Generation Feature Extraction Threshold Generation Decision Similarity Factor YES NO Enrollment Authentication
Swing Your Arm As Can High As You TEXT DEPENDENT 0 1 2 3 4 5 6 7 8 9 x 10 4 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 Time Amplitude
Why?
Voice Ease Of Use Enhanced Security Time Efficient Trust & Confidence
 
 
 
 
 
SYSTEM SPECIFICS Record Training Dialogue Develop Training System Develop Threshold System Develop Decision System  Implement GUI
Record Training Dialogue Develop Training System Develop Threshold System Develop Decision System  Implement GUI
VOICE SAMPLE 1 VOICE SAMPLE 2 VOICE SAMPLE 3
Record Training Dialogue Develop Training System Develop Threshold System Develop Decision System  Implement GUI
 
15-25 ms
FRAME ZCR ENERGY1 ENERGY2 ENERGY3 ENERGY4
VOICE ZCR (1*256) ENERGY1 (1*256) ENERGY2 (1*256) ENERGY3 (1*256) ENERGY4 (1*256)
Speaker-A Euclidean Distance Euclidean Distance VOICE SAMPLE 1 VOICE SAMPLE 2 VOICE SAMPLE 3 ZCR  Energy 1 Energy 2 Energy 3 Energy 4 ZCR  Energy 1 Energy 2 Energy 3 Energy 4 ZCR  Energy 1 Energy 2 Energy 3 Energy 4
Sample  1 & 2 Sample 2 & 3 Sample 3 & 1 Comparison Matrix 1  (5 * 256) Comparison Matrix 2  (5 * 256) Comparison Matrix 3 (5 * 256)
Record Training Dialogue Develop Training System Develop Threshold System Develop Decision System  Implement GUI
Energy1 matrix Energy1 (1) Energy1 (2) Energy1 (3) Energy2 matrix Energy2 (1) Energy2 (2) Energy2 (3) ZCR matrix ZCR1 ZCR2 ZCR3
Energy4 matrix Energy4 (1) Energy4 (2) Energy4 (3) Energy3 matrix Energy3 (1) Energy3 (2) Energy3 (3)
THRESHOLD MATRIX ZCR  Energy 1 Energy 2 Energy 3 Energy 4 MAX MIN
Record Training Dialogue Develop Training System Develop Threshold System Develop Decision System  Implement GUI
I am Karan! Claimant Database of enrolled users Claims to be Karan System prompts claimant to record Claimant gives a speech reference
Claimant’s Speech Feature Extraction
Feature Matrix Sample 1 Feature Matrix Sample 2 Feature Matrix Sample 3 Feature Matrix Euclidean Distance Claimant Claimed Speaker
Comparison Matrix 1  (5 * 256) Comparison Matrix 2  (5 * 256) Comparison Matrix 3 (5 * 256) Threshold Matrix [5*2] Claimant
Threshold Matrix [5*2] Threshold Matrix [5*2] Decision +/- 20% Verified Imposter YES NO Claimant Claimed Speaker
Record Training Dialogue Develop Training System Develop Threshold System Develop Decision System  Implement GUI
 
 
 
TEST PLAN
 
 
 
 
CONCLUSION Working System GUI
PROPOSED ENHANCEMENTS More Voice Features Independent System
ACKNOWLEDGEMENTS Prof. K. Gopalan Prof. Ed Pierson Prof. Don Gray Xia Huang Ankit Toshniwal Devanshu Singh Jitika Bharaj
 

More Related Content

PPT
Speech recognition
PPTX
Text Prompted Remote Speaker Authentication : Joint Speech and Speaker Recogn...
PPTX
PPTX
Speaker Recognition using Gaussian Mixture Model
PPTX
Speaker recognition using MFCC
PPTX
Speech recognition system seminar
PPTX
Speech recognition final presentation
PPTX
Speech recognition An overview
Speech recognition
Text Prompted Remote Speaker Authentication : Joint Speech and Speaker Recogn...
Speaker Recognition using Gaussian Mixture Model
Speaker recognition using MFCC
Speech recognition system seminar
Speech recognition final presentation
Speech recognition An overview

Similar to Speaker Verification System (20)

PPTX
Speaker Recognition
PPTX
Speaker Identification and Verification
PPTX
SPEKER RECOGNITION UNDER LIMITED DATA CODITION
PPT
Automatic speech recognition
PPT
Text Independent Speaker recognitom framework for detecting criminals.ppt
PPTX
Text independent speaker recognition system
PDF
V041203124126
PPTX
DEVELOPMENT OF SPEAKER VERIFICATION UNDER LIMITED DATA AND CONDITION
PPTX
SPEECH RECOGNITION USING NEURAL NETWORK
PDF
ASR_final
PDF
American sign language recognizer
PDF
Bachelors project summary
PDF
Dy36749754
PPTX
SPEAKER VERIFICATION
PDF
B.Tech Project Report
PDF
Final thesis
PPTX
Speaker identification
PDF
IRJET- Voice Command Execution with Speech Recognition and Synthesizer
PPTX
Speaker recognition system by abhishek mahajan
PDF
Utterance based speaker identification
Speaker Recognition
Speaker Identification and Verification
SPEKER RECOGNITION UNDER LIMITED DATA CODITION
Automatic speech recognition
Text Independent Speaker recognitom framework for detecting criminals.ppt
Text independent speaker recognition system
V041203124126
DEVELOPMENT OF SPEAKER VERIFICATION UNDER LIMITED DATA AND CONDITION
SPEECH RECOGNITION USING NEURAL NETWORK
ASR_final
American sign language recognizer
Bachelors project summary
Dy36749754
SPEAKER VERIFICATION
B.Tech Project Report
Final thesis
Speaker identification
IRJET- Voice Command Execution with Speech Recognition and Synthesizer
Speaker recognition system by abhishek mahajan
Utterance based speaker identification
Ad

Recently uploaded (20)

PPT
What is a Computer? Input Devices /output devices
PDF
NewMind AI Weekly Chronicles – August ’25 Week III
PDF
The influence of sentiment analysis in enhancing early warning system model f...
PDF
Produktkatalog für HOBO Datenlogger, Wetterstationen, Sensoren, Software und ...
PDF
sustainability-14-14877-v2.pddhzftheheeeee
PDF
1 - Historical Antecedents, Social Consideration.pdf
PPTX
Microsoft Excel 365/2024 Beginner's training
PDF
Getting started with AI Agents and Multi-Agent Systems
PPT
Geologic Time for studying geology for geologist
PDF
Convolutional neural network based encoder-decoder for efficient real-time ob...
PPTX
The various Industrial Revolutions .pptx
PPT
Galois Field Theory of Risk: A Perspective, Protocol, and Mathematical Backgr...
PPTX
AI IN MARKETING- PRESENTED BY ANWAR KABIR 1st June 2025.pptx
PPTX
Modernising the Digital Integration Hub
PPTX
Configure Apache Mutual Authentication
PDF
A contest of sentiment analysis: k-nearest neighbor versus neural network
PDF
CloudStack 4.21: First Look Webinar slides
DOCX
search engine optimization ppt fir known well about this
PDF
A review of recent deep learning applications in wood surface defect identifi...
PDF
ENT215_Completing-a-large-scale-migration-and-modernization-with-AWS.pdf
What is a Computer? Input Devices /output devices
NewMind AI Weekly Chronicles – August ’25 Week III
The influence of sentiment analysis in enhancing early warning system model f...
Produktkatalog für HOBO Datenlogger, Wetterstationen, Sensoren, Software und ...
sustainability-14-14877-v2.pddhzftheheeeee
1 - Historical Antecedents, Social Consideration.pdf
Microsoft Excel 365/2024 Beginner's training
Getting started with AI Agents and Multi-Agent Systems
Geologic Time for studying geology for geologist
Convolutional neural network based encoder-decoder for efficient real-time ob...
The various Industrial Revolutions .pptx
Galois Field Theory of Risk: A Perspective, Protocol, and Mathematical Backgr...
AI IN MARKETING- PRESENTED BY ANWAR KABIR 1st June 2025.pptx
Modernising the Digital Integration Hub
Configure Apache Mutual Authentication
A contest of sentiment analysis: k-nearest neighbor versus neural network
CloudStack 4.21: First Look Webinar slides
search engine optimization ppt fir known well about this
A review of recent deep learning applications in wood surface defect identifi...
ENT215_Completing-a-large-scale-migration-and-modernization-with-AWS.pdf
Ad

Speaker Verification System