building a robust facial recognition system based … presentation... · building a robust facial...

24
End of Year Presentation Principle Investigator: David Pilkington Supervisor: James Connan BUILDING A ROBUST FACIAL RECOGNITION SYSTEM BASED ON GENERIC TOOLS

Upload: truongtruc

Post on 29-Aug-2018

220 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

End of Year Presentation

Principle Investigator: David Pilkington

Supervisor: James Connan

BUILDING A ROBUST FACIAL RECOGNITION SYSTEM BASED ON GENERIC TOOLS

Page 2: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

CONTENTS • Introduction

• Recap

• Changes to image database

• Testing Framework

• Benchmark System

• Design

• Testing Results

• Weaknesses and Proposed Solutions

• Developed System

• Design

• Implementation and Encountered Problems

• Experimental Results

• Concluding Thoughts and Possible Future Work

2

Page 3: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

RECAP • Makes use of the EmguCV package (C# wrapper for OpenCV)

• Logitech 300 webcam (640 x 480)

• Benchmark System to establish baseline performance

• Identify weaknesses

• Propose and implement solutions

3

Page 4: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

CHANGES TO THE IMAGE DATABASE • Comprised of 50 images

• 10 individuals with 5 images each

• Captured using the tools provided by the package

• Stored as .png files due to lossless compression (640 x 480)

• Varying backgrounds, lighting conditions, sexes, race

4

Page 5: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

TESTING FRAMEWORK

5

• Consists of 5 tests:

Test Description 1 1 training image 2 2 training images 3 3 training images 4 4 training images 5 5 training images

Page 6: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

BENCHMARK SYSTEM

6

Page 7: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

DESIGN

Input Image EigenObjectRe

cogniser Identified

Face

7

training

Page 8: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

TEST RESULTS

8

Test Recognition Rate 1 40% (s2, s3) 2 40% (s2, s3) 3 40% (s2, s3) 4 40% (s2, s3) 5 40% (s2, s3)

Page 9: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

IDENTIFIED WEAKNESSES

• Lower recognition rate than we would like (40%)

• No ability to reject an image

• Highly afftected by spurious background noise

• Highly affected by varying lighting

9

Page 10: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

PROPOSED SOLUTION

10

• Make use of face detection to crop image

• Make use of skin segmentation to further reduce background noise

• Use histogram equalisation to reduce light variability

• Introduce a threshold to the recongniser to allow the rejection of images

Page 11: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

DEVELOPED SYSTEM

11

Page 12: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

DESIGN

Input Image Face Detector Skin Segmentation

Eigen Recogniser

Identified Face

Image Rejected

12

training

Page 13: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

FACE DETECTOR

13

Page 14: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

SKIN SEGMENTATION: DESIGN

14

Input Image Nose Detector Colour Modelling Segmentation Segmented

Image

Page 15: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

SKIN SEGMENTOR : NOSE IDENTIFICATION

15

Page 16: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

SKIN SEGMENTOR: SKIN MODELING

16

Page 17: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

SKIN SEGMENTOR: SEGMENTATION

17

Page 18: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

EIGEN RECOGNISER

18

TRAINING

• Makes use of the eigenfaces technique which is base on PCA

• Receive training images and labels

• Create the PCA subspace

• Compute eigenvectors for training images

• Project the training images onto the PCA subspace to obtain eigenvalues for images

RECOGNITION

• Project test image onto the PCA subspace and obtain eigenvalues

• Find nearest neighbour by shortest Euclidean Distance

Page 19: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

EXPERIMENTAL RESULTS

19

Test Recognition Rate 1 60% (s1, s2, s4) 2 60% (s1, s2, s4) 3 60%(s1, s2, s4) 4 80% (s1, s2, s3, s4) 5 80% (s1, s2, s3, s4)

Page 20: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

EXPERIMENTAL RESULTS : COMPARISON

Euclidean Distance to Nearest Neighbour

20

Page 21: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

EXPERIMENTAL RESULTS : COMPARISON

21

VS

Benchmark System Developed System

Page 22: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

EXPERIMENTAL RESULTS : COMPARISON

22

Database Load Time

Page 23: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

CONCLUDING THOUGHTS AND POSSIBLE FUTURE WORK

Conclusions • Background noise reduction has proved to be effective • Light sensitivity is still a problem • Thresholding was temperamental • Scalability problem due to image processes

Possible Future Extensions • Expand image data set • Comparison between recognition techniques • Effects of glasses, beards, etc. • Image verification

23

Page 24: Building a robust facial recognition system based … Presentation... · BUILDING A ROBUST FACIAL RECOGNITION ... • Make use of face detection to crop image ... Building a robust

PROUDLY BROUGHT TO YOU BY …

24