SlideShare a Scribd company logo
Gait Based Authentication
System
1
Project Guide
Mr. Vikas Mittal,
(HOD CS & E)
Project Members
Vatsal Agarwal
Vivek Kumar
Vijay Bhasin
Yash Agarwal
About Project
It aims to authenticate subject on the basis of one’s gait - i.e. method of
walking.
Everyone has it’s own unique gait - an important biometric.
Clinical research has shown that gait, the way a person walks or runs, is
sufficient to distinguish between individuals [1].
Can be thought of as unintrusive authentication method as one can easily get
authenticated without using tedious challenges as username - password /
PIN combination or fingerprint biometric. 2
About contd ..
Authentication is a specialized case of identification.
Aim to set a new standard in mobile security.
Project Site - bit.ly/gaitAuth
3
Scope
● High end security solution
● Effective as uses minimal / no user interaction
● Can be used in security intensive arenas as :
○ High security compounds
○ Logging entry / exit
○ Access protected areas
4
Implementation
Types of implementation possible
a. Machine Vision Technology (MVT) [3,4]
b. Floor Sensor Technology (FST) [5]
c. Wearable Sensor Technology (WST) [6]
WST is recognised as most approachable and newest of all.
Sensors in WST are attached to human body in various positions as pockets,
waist or shoes to record physical motions. 5
Implementation contd ..
Phases of Implementation
Gathering Data
Feature Extraction and Data Transformation
Modelling
Training
Testing
6
7
Gathering Data
3 axes accelerometer
data is logged into
csv format by an
android application [2]
Dedicated portal is
designed to
encourage
participation and data
upload facility is
provided [7] .
8
Feature Extraction and Data Transformation
Gait features are extracted to compare between genuine and imposter
attempt.
Transformation process partitions the time series data into a fixed time non
overlapping windows and generates relatively simple features based on
data in window [8] .
9
Feature Extraction contd ..
Features to be mined :
Average - Average sensor value each axis
Standard Deviation - SD for each axis
Average Absolute Difference - AAD between set of values and mean of these values
Time Between Peaks - Between sinusoidal waves
Binned Distribution - Range of values determined, x equal sized bins formed
Average Resultant Acceleration - For each of the sensor values in window, take square root 10
Modelling
Classifier Induction Algorithms are used.
The WEKA data mining suite is freely available and implements a large
number of classifier algorithms.
The algorithms that can prove effective for gait are :
Multilayer Perceptron (MLP) - Neural Network Based
Random Forest
Rotation Forest 11
Training
The task is to identify a user from pool of users based on sample of their
sample data.
Requires sample of data from all users in the population to use in training.
The transformed datasets associated with accelerometer data are used to
train and evaluate biometric identification models.
10 fold cross validation is used.
12
Testing
Testing is done on the left datasets using the classifier algorithms.
Random users are picked and any x minute of data is chosen and classifier
model runs on the data to identify a genuine or imposter attempt.
The process is repeated to get accurate results.
13
Validating
In authentication each user has their own model and when a sample of
sensor data is provided, the task is to determine if the sample belongs to
user or to an imposter.
The results are based on identifying each user based on a single x minutes
data sample(instance).
Overall accuracy will be the total number of correct predictions divided by
total number of predictions.
14
GBAS Portal
15
Upload Facility
16
GBAS Portal Hits
17
References
1. M. Nixon, T. Tan, and R. Chellapppa, Human Identification Based on Gait.
New York: Springer Science + Business Media Inc., 2006, ch. 1.
2. Accelerometer Acceleration Log
https://play.google.com/store/apps/details?id=com.alfav.applications.acc
elerometerlog
3. L. Lee, “Gait Analysis for classification”, Massachusetts Institute of
Technology-artificial Intelligence Laboratory, 2003.
18
References contd ..
19
4. L. Wang, T. Tan, W. Hu and H. Ning, “Automatic gait recognition based on
statistical shape analysis”, in IEEE Transactions on Image Processing, 2003.
5. J. Suutala, K. Fujinami and J. Roning, “Gaussian process person identifier based
on simple floor sensor”, in Proceedings of the 3rd European Conference on Smart
Sensing and Context, 2008.
6. A. Annadhorai, E. Guenterberg, J. Barnes and K. Haraga, “Human identification by
gait analysis”, in Proceedings of the 2nd International Workshop on Systems and
Networking Support for Health Care and Assisted Living Environments, 2008.
7. GBAS portal - bit.ly/gaitAuth, upload facility - bit.ly/gaitData
8. R windows - https://stat.ethz.ch/R-manual/R-
devel/library/stats/html/window.html
Contributors
1. Arpit Agarwal, Greater Noida
2. Shubhanshu Arora, Moradabad
3. Rajat Agarwal, Bangalore
4. Sugam, Moradabad
5. Vatsal, Vivek , Vijay, Yash - Moradabad
20

More Related Content

What's hot

APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...
APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...
APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...
IAEME Publication
 
AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)
AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)
AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)
csandit
 
A-Z of AI in Radiology
A-Z of AI in RadiologyA-Z of AI in Radiology
A-Z of AI in Radiology
Dr Hugh Harvey
 
Data science lecture4_doaa_mohey
Data science lecture4_doaa_moheyData science lecture4_doaa_mohey
Data science lecture4_doaa_mohey
Doaa Mohey Eldin
 
A comparative analysis of data mining tools for performance mapping of wlan data
A comparative analysis of data mining tools for performance mapping of wlan dataA comparative analysis of data mining tools for performance mapping of wlan data
A comparative analysis of data mining tools for performance mapping of wlan dataIAEME Publication
 
Concept Drift Identification using Classifier Ensemble Approach
Concept Drift Identification using Classifier Ensemble Approach  Concept Drift Identification using Classifier Ensemble Approach
Concept Drift Identification using Classifier Ensemble Approach
IJECEIAES
 
Homework 9 17-2011
Homework 9 17-2011Homework 9 17-2011
Homework 9 17-2011
d015125
 
Gender Classification based on Mouse Movement Activity on Computer
Gender Classification based on Mouse Movement Activity on ComputerGender Classification based on Mouse Movement Activity on Computer
Gender Classification based on Mouse Movement Activity on Computer
Amin Ismail
 
Thesis presentation ist
Thesis presentation istThesis presentation ist
Thesis presentation ist
deep sharma
 
Performance Comparision of Machine Learning Algorithms
Performance Comparision of Machine Learning AlgorithmsPerformance Comparision of Machine Learning Algorithms
Performance Comparision of Machine Learning Algorithms
Dinusha Dilanka
 
Disease Identification and Detection in Apple Tree
Disease Identification and Detection in Apple TreeDisease Identification and Detection in Apple Tree
Disease Identification and Detection in Apple Tree
ijtsrd
 
Comparative Analysis: Effective Information Retrieval Using Different Learnin...
Comparative Analysis: Effective Information Retrieval Using Different Learnin...Comparative Analysis: Effective Information Retrieval Using Different Learnin...
Comparative Analysis: Effective Information Retrieval Using Different Learnin...
RSIS International
 
Data repository for sensor network a data mining approach
Data repository for sensor network  a data mining approachData repository for sensor network  a data mining approach
Data repository for sensor network a data mining approach
ijdms
 
Fault Detection in Mobile Communication Networks Using Data Mining Techniques...
Fault Detection in Mobile Communication Networks Using Data Mining Techniques...Fault Detection in Mobile Communication Networks Using Data Mining Techniques...
Fault Detection in Mobile Communication Networks Using Data Mining Techniques...
ijcisjournal
 
Fi cloudpresentationgyrardaugust2015 v2
Fi cloudpresentationgyrardaugust2015 v2Fi cloudpresentationgyrardaugust2015 v2
Fi cloudpresentationgyrardaugust2015 v2
Amélie Gyrard
 
The International Journal of Engineering and Science (The IJES)
The International Journal of Engineering and Science (The IJES)The International Journal of Engineering and Science (The IJES)
The International Journal of Engineering and Science (The IJES)
theijes
 
Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...
Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...
Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...
Mahmud Hossain
 

What's hot (17)

APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...
APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...
APPLICATION OF SELF-ORGANIZING FEATURE MAPS AND MARKOV CHAINS TO RECOGNITION ...
 
AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)
AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)
AUTOMATIC THEFT SECURITY SYSTEM (SMART SURVEILLANCE CAMERA)
 
A-Z of AI in Radiology
A-Z of AI in RadiologyA-Z of AI in Radiology
A-Z of AI in Radiology
 
Data science lecture4_doaa_mohey
Data science lecture4_doaa_moheyData science lecture4_doaa_mohey
Data science lecture4_doaa_mohey
 
A comparative analysis of data mining tools for performance mapping of wlan data
A comparative analysis of data mining tools for performance mapping of wlan dataA comparative analysis of data mining tools for performance mapping of wlan data
A comparative analysis of data mining tools for performance mapping of wlan data
 
Concept Drift Identification using Classifier Ensemble Approach
Concept Drift Identification using Classifier Ensemble Approach  Concept Drift Identification using Classifier Ensemble Approach
Concept Drift Identification using Classifier Ensemble Approach
 
Homework 9 17-2011
Homework 9 17-2011Homework 9 17-2011
Homework 9 17-2011
 
Gender Classification based on Mouse Movement Activity on Computer
Gender Classification based on Mouse Movement Activity on ComputerGender Classification based on Mouse Movement Activity on Computer
Gender Classification based on Mouse Movement Activity on Computer
 
Thesis presentation ist
Thesis presentation istThesis presentation ist
Thesis presentation ist
 
Performance Comparision of Machine Learning Algorithms
Performance Comparision of Machine Learning AlgorithmsPerformance Comparision of Machine Learning Algorithms
Performance Comparision of Machine Learning Algorithms
 
Disease Identification and Detection in Apple Tree
Disease Identification and Detection in Apple TreeDisease Identification and Detection in Apple Tree
Disease Identification and Detection in Apple Tree
 
Comparative Analysis: Effective Information Retrieval Using Different Learnin...
Comparative Analysis: Effective Information Retrieval Using Different Learnin...Comparative Analysis: Effective Information Retrieval Using Different Learnin...
Comparative Analysis: Effective Information Retrieval Using Different Learnin...
 
Data repository for sensor network a data mining approach
Data repository for sensor network  a data mining approachData repository for sensor network  a data mining approach
Data repository for sensor network a data mining approach
 
Fault Detection in Mobile Communication Networks Using Data Mining Techniques...
Fault Detection in Mobile Communication Networks Using Data Mining Techniques...Fault Detection in Mobile Communication Networks Using Data Mining Techniques...
Fault Detection in Mobile Communication Networks Using Data Mining Techniques...
 
Fi cloudpresentationgyrardaugust2015 v2
Fi cloudpresentationgyrardaugust2015 v2Fi cloudpresentationgyrardaugust2015 v2
Fi cloudpresentationgyrardaugust2015 v2
 
The International Journal of Engineering and Science (The IJES)
The International Journal of Engineering and Science (The IJES)The International Journal of Engineering and Science (The IJES)
The International Journal of Engineering and Science (The IJES)
 
Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...
Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...
Trust-IoV: A Trustworthy Forensic Investigation Framework for the Internet of...
 

Viewers also liked

FUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICES
FUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICESFUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICES
FUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICESvasim hasina
 
Biometrics Authentication Technology by Sayak Das
Biometrics Authentication Technology by Sayak DasBiometrics Authentication Technology by Sayak Das
Biometrics Authentication Technology by Sayak Das
Tenet Systems Pvt Ltd
 
gait analysis
gait analysisgait analysis
gait analysis
Manoz Kumar
 
Gait analysis and.ppt by ramachandra
Gait analysis and.ppt by ramachandraGait analysis and.ppt by ramachandra
Gait analysis and.ppt by ramachandra
ramachandra reddy
 
Normal and pathological gait
Normal and pathological gaitNormal and pathological gait
Normal and pathological gait
sharanyapalanisshami
 
A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...
A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...
A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...
Michael J. Montgomery
 
Gait ppt
Gait pptGait ppt
Gait analysis
Gait analysisGait analysis
Gait analysis
Aarti Sareen
 
Gait
GaitGait
Slide-show on Biometrics
Slide-show on BiometricsSlide-show on Biometrics
Slide-show on BiometricsPathik504
 
Bio-metrics Technology
Bio-metrics TechnologyBio-metrics Technology
Bio-metrics Technology
Avanitrambadiya
 
Biometric Presentation
Biometric PresentationBiometric Presentation
Biometric Presentationrs2003
 
Biometric's final ppt
Biometric's final pptBiometric's final ppt
Biometric's final ppt
Ankita Vanage
 
Biometric Security advantages and disadvantages
Biometric Security advantages and disadvantagesBiometric Security advantages and disadvantages
Biometric Security advantages and disadvantages
Prabh Jeet
 
Gait normal & abnormal
Gait normal & abnormalGait normal & abnormal
Gait normal & abnormalRatan Khuman
 

Viewers also liked (18)

FUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICES
FUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICESFUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICES
FUSION OF GAIT AND FINGERPRINT FOR USER AUTHENTICATION ON MOBILE DEVICES
 
Kb gait-recognition
Kb gait-recognitionKb gait-recognition
Kb gait-recognition
 
Biometrics Authentication Technology by Sayak Das
Biometrics Authentication Technology by Sayak DasBiometrics Authentication Technology by Sayak Das
Biometrics Authentication Technology by Sayak Das
 
BATCH 11
BATCH 11BATCH 11
BATCH 11
 
gait analysis
gait analysisgait analysis
gait analysis
 
Gait analysis and.ppt by ramachandra
Gait analysis and.ppt by ramachandraGait analysis and.ppt by ramachandra
Gait analysis and.ppt by ramachandra
 
Normal and pathological gait
Normal and pathological gaitNormal and pathological gait
Normal and pathological gait
 
A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...
A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...
A Wearable Accelerometer System for Unobtrusive Monitoring of Parkinson’s Dis...
 
Gait ppt
Gait pptGait ppt
Gait ppt
 
Gait analysis
Gait analysisGait analysis
Gait analysis
 
Gait
GaitGait
Gait
 
Slide-show on Biometrics
Slide-show on BiometricsSlide-show on Biometrics
Slide-show on Biometrics
 
Bio-metrics Technology
Bio-metrics TechnologyBio-metrics Technology
Bio-metrics Technology
 
Biometric Presentation
Biometric PresentationBiometric Presentation
Biometric Presentation
 
Biometric's final ppt
Biometric's final pptBiometric's final ppt
Biometric's final ppt
 
Gait analysis
Gait analysisGait analysis
Gait analysis
 
Biometric Security advantages and disadvantages
Biometric Security advantages and disadvantagesBiometric Security advantages and disadvantages
Biometric Security advantages and disadvantages
 
Gait normal & abnormal
Gait normal & abnormalGait normal & abnormal
Gait normal & abnormal
 

Similar to GaitProjectProposal

Human Activity Recognition Using Neural Network
Human Activity Recognition Using Neural NetworkHuman Activity Recognition Using Neural Network
Human Activity Recognition Using Neural Network
IRJET Journal
 
IRJET- Design an Approach for Prediction of Human Activity Recognition us...
IRJET-  	  Design an Approach for Prediction of Human Activity Recognition us...IRJET-  	  Design an Approach for Prediction of Human Activity Recognition us...
IRJET- Design an Approach for Prediction of Human Activity Recognition us...
IRJET Journal
 
Human Activity Recognition Using Smartphone
Human Activity Recognition Using SmartphoneHuman Activity Recognition Using Smartphone
Human Activity Recognition Using Smartphone
IRJET Journal
 
smartwatch-user-identification
smartwatch-user-identificationsmartwatch-user-identification
smartwatch-user-identificationSebastian W. Cheah
 
IRJET- Surveillance of Object Motion Detection and Caution System using B...
IRJET-  	  Surveillance of Object Motion Detection and Caution System using B...IRJET-  	  Surveillance of Object Motion Detection and Caution System using B...
IRJET- Surveillance of Object Motion Detection and Caution System using B...
IRJET Journal
 
IRJET- Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...
IRJET-  	  Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...IRJET-  	  Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...
IRJET- Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...
IRJET Journal
 
IRJET- Human Activity Recognition using Flex Sensors
IRJET- Human Activity Recognition using Flex SensorsIRJET- Human Activity Recognition using Flex Sensors
IRJET- Human Activity Recognition using Flex Sensors
IRJET Journal
 
A feature selection method based on auto-encoder for internet of things intru...
A feature selection method based on auto-encoder for internet of things intru...A feature selection method based on auto-encoder for internet of things intru...
A feature selection method based on auto-encoder for internet of things intru...
IJECEIAES
 
IoT Based Human Activity Recognition and Classification Using Machine Learning
IoT Based Human Activity Recognition and Classification Using Machine LearningIoT Based Human Activity Recognition and Classification Using Machine Learning
IoT Based Human Activity Recognition and Classification Using Machine Learning
IRJET Journal
 
Network Intrusion Detection System using Machine Learning
Network Intrusion Detection System using Machine LearningNetwork Intrusion Detection System using Machine Learning
Network Intrusion Detection System using Machine Learning
IRJET Journal
 
Online Signature Authentication by Using Mouse Behavior
Online Signature Authentication by Using Mouse Behavior Online Signature Authentication by Using Mouse Behavior
Online Signature Authentication by Using Mouse Behavior
Editor IJCATR
 
IRJET- FASSBTR : Fingerprint Authentication System Security using Barcode...
IRJET-  	  FASSBTR : Fingerprint Authentication System Security using Barcode...IRJET-  	  FASSBTR : Fingerprint Authentication System Security using Barcode...
IRJET- FASSBTR : Fingerprint Authentication System Security using Barcode...
IRJET Journal
 
H364752
H364752H364752
H364752
IJERA Editor
 
IRJET- Review on Human Action Detection in Stored Videos using Support Vector...
IRJET- Review on Human Action Detection in Stored Videos using Support Vector...IRJET- Review on Human Action Detection in Stored Videos using Support Vector...
IRJET- Review on Human Action Detection in Stored Videos using Support Vector...
IRJET Journal
 
Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...
Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...
Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...
Drjabez
 
Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...
Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...
Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...
AIRCC Publishing Corporation
 
IRJET- Anomaly Detection System in CCTV Derived Videos
IRJET- Anomaly Detection System in CCTV Derived VideosIRJET- Anomaly Detection System in CCTV Derived Videos
IRJET- Anomaly Detection System in CCTV Derived Videos
IRJET Journal
 
Draft activity recognition from accelerometer data
Draft activity recognition from accelerometer dataDraft activity recognition from accelerometer data
Draft activity recognition from accelerometer data
Raghu Palakodety
 
Review on Object Counting System
Review on Object Counting SystemReview on Object Counting System
Review on Object Counting System
IRJET Journal
 
IRJET- Criminal Recognization in CCTV Surveillance Video
IRJET-  	  Criminal Recognization in CCTV Surveillance VideoIRJET-  	  Criminal Recognization in CCTV Surveillance Video
IRJET- Criminal Recognization in CCTV Surveillance Video
IRJET Journal
 

Similar to GaitProjectProposal (20)

Human Activity Recognition Using Neural Network
Human Activity Recognition Using Neural NetworkHuman Activity Recognition Using Neural Network
Human Activity Recognition Using Neural Network
 
IRJET- Design an Approach for Prediction of Human Activity Recognition us...
IRJET-  	  Design an Approach for Prediction of Human Activity Recognition us...IRJET-  	  Design an Approach for Prediction of Human Activity Recognition us...
IRJET- Design an Approach for Prediction of Human Activity Recognition us...
 
Human Activity Recognition Using Smartphone
Human Activity Recognition Using SmartphoneHuman Activity Recognition Using Smartphone
Human Activity Recognition Using Smartphone
 
smartwatch-user-identification
smartwatch-user-identificationsmartwatch-user-identification
smartwatch-user-identification
 
IRJET- Surveillance of Object Motion Detection and Caution System using B...
IRJET-  	  Surveillance of Object Motion Detection and Caution System using B...IRJET-  	  Surveillance of Object Motion Detection and Caution System using B...
IRJET- Surveillance of Object Motion Detection and Caution System using B...
 
IRJET- Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...
IRJET-  	  Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...IRJET-  	  Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...
IRJET- Survey on Shoulder Surfing Resistant Pin Entry by using Base Pin a...
 
IRJET- Human Activity Recognition using Flex Sensors
IRJET- Human Activity Recognition using Flex SensorsIRJET- Human Activity Recognition using Flex Sensors
IRJET- Human Activity Recognition using Flex Sensors
 
A feature selection method based on auto-encoder for internet of things intru...
A feature selection method based on auto-encoder for internet of things intru...A feature selection method based on auto-encoder for internet of things intru...
A feature selection method based on auto-encoder for internet of things intru...
 
IoT Based Human Activity Recognition and Classification Using Machine Learning
IoT Based Human Activity Recognition and Classification Using Machine LearningIoT Based Human Activity Recognition and Classification Using Machine Learning
IoT Based Human Activity Recognition and Classification Using Machine Learning
 
Network Intrusion Detection System using Machine Learning
Network Intrusion Detection System using Machine LearningNetwork Intrusion Detection System using Machine Learning
Network Intrusion Detection System using Machine Learning
 
Online Signature Authentication by Using Mouse Behavior
Online Signature Authentication by Using Mouse Behavior Online Signature Authentication by Using Mouse Behavior
Online Signature Authentication by Using Mouse Behavior
 
IRJET- FASSBTR : Fingerprint Authentication System Security using Barcode...
IRJET-  	  FASSBTR : Fingerprint Authentication System Security using Barcode...IRJET-  	  FASSBTR : Fingerprint Authentication System Security using Barcode...
IRJET- FASSBTR : Fingerprint Authentication System Security using Barcode...
 
H364752
H364752H364752
H364752
 
IRJET- Review on Human Action Detection in Stored Videos using Support Vector...
IRJET- Review on Human Action Detection in Stored Videos using Support Vector...IRJET- Review on Human Action Detection in Stored Videos using Support Vector...
IRJET- Review on Human Action Detection in Stored Videos using Support Vector...
 
Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...
Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...
Intrusion Detection System (IDS): Anomaly Detection using Outlier Detection A...
 
Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...
Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...
Your Device May Know You Better Than You Know Yourself-Continuous Authenticat...
 
IRJET- Anomaly Detection System in CCTV Derived Videos
IRJET- Anomaly Detection System in CCTV Derived VideosIRJET- Anomaly Detection System in CCTV Derived Videos
IRJET- Anomaly Detection System in CCTV Derived Videos
 
Draft activity recognition from accelerometer data
Draft activity recognition from accelerometer dataDraft activity recognition from accelerometer data
Draft activity recognition from accelerometer data
 
Review on Object Counting System
Review on Object Counting SystemReview on Object Counting System
Review on Object Counting System
 
IRJET- Criminal Recognization in CCTV Surveillance Video
IRJET-  	  Criminal Recognization in CCTV Surveillance VideoIRJET-  	  Criminal Recognization in CCTV Surveillance Video
IRJET- Criminal Recognization in CCTV Surveillance Video
 

GaitProjectProposal

  • 1. Gait Based Authentication System 1 Project Guide Mr. Vikas Mittal, (HOD CS & E) Project Members Vatsal Agarwal Vivek Kumar Vijay Bhasin Yash Agarwal
  • 2. About Project It aims to authenticate subject on the basis of one’s gait - i.e. method of walking. Everyone has it’s own unique gait - an important biometric. Clinical research has shown that gait, the way a person walks or runs, is sufficient to distinguish between individuals [1]. Can be thought of as unintrusive authentication method as one can easily get authenticated without using tedious challenges as username - password / PIN combination or fingerprint biometric. 2
  • 3. About contd .. Authentication is a specialized case of identification. Aim to set a new standard in mobile security. Project Site - bit.ly/gaitAuth 3
  • 4. Scope ● High end security solution ● Effective as uses minimal / no user interaction ● Can be used in security intensive arenas as : ○ High security compounds ○ Logging entry / exit ○ Access protected areas 4
  • 5. Implementation Types of implementation possible a. Machine Vision Technology (MVT) [3,4] b. Floor Sensor Technology (FST) [5] c. Wearable Sensor Technology (WST) [6] WST is recognised as most approachable and newest of all. Sensors in WST are attached to human body in various positions as pockets, waist or shoes to record physical motions. 5
  • 6. Implementation contd .. Phases of Implementation Gathering Data Feature Extraction and Data Transformation Modelling Training Testing 6
  • 7. 7
  • 8. Gathering Data 3 axes accelerometer data is logged into csv format by an android application [2] Dedicated portal is designed to encourage participation and data upload facility is provided [7] . 8
  • 9. Feature Extraction and Data Transformation Gait features are extracted to compare between genuine and imposter attempt. Transformation process partitions the time series data into a fixed time non overlapping windows and generates relatively simple features based on data in window [8] . 9
  • 10. Feature Extraction contd .. Features to be mined : Average - Average sensor value each axis Standard Deviation - SD for each axis Average Absolute Difference - AAD between set of values and mean of these values Time Between Peaks - Between sinusoidal waves Binned Distribution - Range of values determined, x equal sized bins formed Average Resultant Acceleration - For each of the sensor values in window, take square root 10
  • 11. Modelling Classifier Induction Algorithms are used. The WEKA data mining suite is freely available and implements a large number of classifier algorithms. The algorithms that can prove effective for gait are : Multilayer Perceptron (MLP) - Neural Network Based Random Forest Rotation Forest 11
  • 12. Training The task is to identify a user from pool of users based on sample of their sample data. Requires sample of data from all users in the population to use in training. The transformed datasets associated with accelerometer data are used to train and evaluate biometric identification models. 10 fold cross validation is used. 12
  • 13. Testing Testing is done on the left datasets using the classifier algorithms. Random users are picked and any x minute of data is chosen and classifier model runs on the data to identify a genuine or imposter attempt. The process is repeated to get accurate results. 13
  • 14. Validating In authentication each user has their own model and when a sample of sensor data is provided, the task is to determine if the sample belongs to user or to an imposter. The results are based on identifying each user based on a single x minutes data sample(instance). Overall accuracy will be the total number of correct predictions divided by total number of predictions. 14
  • 18. References 1. M. Nixon, T. Tan, and R. Chellapppa, Human Identification Based on Gait. New York: Springer Science + Business Media Inc., 2006, ch. 1. 2. Accelerometer Acceleration Log https://play.google.com/store/apps/details?id=com.alfav.applications.acc elerometerlog 3. L. Lee, “Gait Analysis for classification”, Massachusetts Institute of Technology-artificial Intelligence Laboratory, 2003. 18
  • 19. References contd .. 19 4. L. Wang, T. Tan, W. Hu and H. Ning, “Automatic gait recognition based on statistical shape analysis”, in IEEE Transactions on Image Processing, 2003. 5. J. Suutala, K. Fujinami and J. Roning, “Gaussian process person identifier based on simple floor sensor”, in Proceedings of the 3rd European Conference on Smart Sensing and Context, 2008. 6. A. Annadhorai, E. Guenterberg, J. Barnes and K. Haraga, “Human identification by gait analysis”, in Proceedings of the 2nd International Workshop on Systems and Networking Support for Health Care and Assisted Living Environments, 2008. 7. GBAS portal - bit.ly/gaitAuth, upload facility - bit.ly/gaitData 8. R windows - https://stat.ethz.ch/R-manual/R- devel/library/stats/html/window.html
  • 20. Contributors 1. Arpit Agarwal, Greater Noida 2. Shubhanshu Arora, Moradabad 3. Rajat Agarwal, Bangalore 4. Sugam, Moradabad 5. Vatsal, Vivek , Vijay, Yash - Moradabad 20

Editor's Notes

  1. SVM - hyper plane Naive - Cause relationship Multilayer perceptron - definite i/p - o/p multilayer composed of nodes, each layer connected Random forest - decision trees, reduces overfitting Rotation Forest - How and how many classes should be eliminated, uses PCA - you can avoid occasionally identical selection of features across the L classifiers
  2. Data is fed into the algorithm and coefficients and output is validated to iterate or continue.