PPT On Face Recognition Techniques
Download
Face Recognition Techniques Presentation Transcript:
1.A Review On Face Recognition Techniques
2.Contents
Introduction
Review on the topic
Related Issues
Face Recognition Techniques
Comparison
Discussion
Databases
Conclusion
References
3.Introduction
Face recognition is a biometric technique for automatic identification.
It makes use the most usual human identifier, face and unique facial characteristics.
Emergence has been fuelled due to wide range of law and commercial enforcement.
Widely used in commercial systems to perform real-time face detection, image registration and image matching.
4.Technologies in face recognition have been evolving through years. Its importance has recently grown in a significant manner due to:
Increased civilian and commercial research projects.
Need for surveillance in trafficking.
Increased terrorist activities.
Enhanced real time computation and exploration of real time hardware.
5. Review
Before the middle 90’s, the research attention was only focused on single-face recognition. The approaches included:
Kirby and Sirovich [1] were among the first to apply principal component analysis (PCA) to face images.
Turk and Pentland popularized PCA for face recognition [4] via eigen faces.
Neural networks have been widely applied in pattern recognition as achieve better performance than the simple minimum distance classifiers [16].
Using skin color etc.
6.During the past ten years, considerable progress has been made in multi-face recognition area, includes: RBF neural networks process implemented in helps in structure determination of the radial basis function (RBF) neural networks.
Discrete cosine transform (DCT) [3, 14].
Support vector machine (SVM) by Osuna et al. (1997).
Hidden Markov Model [19, 39].
Multilinear PCA [31, 32].
Face Recognition using Texture and Depth Information[38].
7.Technical approaches to face recognition
Feature based approach : based on shape and geometrical relationships of key facial features including eyes, mouth, nose, chin and curvature based face components [18]. These are more robust against rotation, scale, and illumination variations
Holistic approach (Template matching approach) : takes the input face images globally and extract important facial features based on the high-dimensional intensity values of face images automatically. They greatly rely on the accuracy of facial feature detection
Hybrid approach : uses both the face images together with the local features for face recognition.
8.Issues faced by face recognition techniques
Image intensity and orientation
Pose
Structural components
Occlusion
Image quality
Facial expression
Illumination
9.Face Recognition Techniques
The incredible human intelligence can be demonstrated by its ability to recognize human faces.
Over the last three decades researches have been going on to study this outstanding visual perception of human beings in machine recognition of faces.
While coping up with the challenges in face recognition numerous techniques have been implemented and few are as follows
10.Principal component analysis (PCA) is a statistical dimensionality reduction method.
This subspace projection technique has found application in fields such as face recognition, pattern recognition and image compression.
It is computationally efficient to compare images in subspaces with significantly reduced dimensions.
PCA helps to reduce image vectors with 65,536 pixels (256x256) might be projected into a subspace with only 100 to 300 dimensions.
PCA reveals the most effective low dimensional structure of facial patterns by decomposing the face structure into orthogonal (uncorrelated) components known as eigenvectors and eigenvalues [20].
Download
Face Recognition Techniques Presentation Transcript:
1.A Review On Face Recognition Techniques
2.Contents
Introduction
Review on the topic
Related Issues
Face Recognition Techniques
Comparison
Discussion
Databases
Conclusion
References
3.Introduction
Face recognition is a biometric technique for automatic identification.
It makes use the most usual human identifier, face and unique facial characteristics.
Emergence has been fuelled due to wide range of law and commercial enforcement.
Widely used in commercial systems to perform real-time face detection, image registration and image matching.
4.Technologies in face recognition have been evolving through years. Its importance has recently grown in a significant manner due to:
Increased civilian and commercial research projects.
Need for surveillance in trafficking.
Increased terrorist activities.
Enhanced real time computation and exploration of real time hardware.
5. Review
Before the middle 90’s, the research attention was only focused on single-face recognition. The approaches included:
Kirby and Sirovich [1] were among the first to apply principal component analysis (PCA) to face images.
Turk and Pentland popularized PCA for face recognition [4] via eigen faces.
Neural networks have been widely applied in pattern recognition as achieve better performance than the simple minimum distance classifiers [16].
Using skin color etc.
6.During the past ten years, considerable progress has been made in multi-face recognition area, includes: RBF neural networks process implemented in helps in structure determination of the radial basis function (RBF) neural networks.
Discrete cosine transform (DCT) [3, 14].
Support vector machine (SVM) by Osuna et al. (1997).
Hidden Markov Model [19, 39].
Multilinear PCA [31, 32].
Face Recognition using Texture and Depth Information[38].
7.Technical approaches to face recognition
Feature based approach : based on shape and geometrical relationships of key facial features including eyes, mouth, nose, chin and curvature based face components [18]. These are more robust against rotation, scale, and illumination variations
Holistic approach (Template matching approach) : takes the input face images globally and extract important facial features based on the high-dimensional intensity values of face images automatically. They greatly rely on the accuracy of facial feature detection
Hybrid approach : uses both the face images together with the local features for face recognition.
8.Issues faced by face recognition techniques
Image intensity and orientation
Pose
Structural components
Occlusion
Image quality
Facial expression
Illumination
9.Face Recognition Techniques
The incredible human intelligence can be demonstrated by its ability to recognize human faces.
Over the last three decades researches have been going on to study this outstanding visual perception of human beings in machine recognition of faces.
While coping up with the challenges in face recognition numerous techniques have been implemented and few are as follows
10.Principal component analysis (PCA) is a statistical dimensionality reduction method.
This subspace projection technique has found application in fields such as face recognition, pattern recognition and image compression.
It is computationally efficient to compare images in subspaces with significantly reduced dimensions.
PCA helps to reduce image vectors with 65,536 pixels (256x256) might be projected into a subspace with only 100 to 300 dimensions.
PCA reveals the most effective low dimensional structure of facial patterns by decomposing the face structure into orthogonal (uncorrelated) components known as eigenvectors and eigenvalues [20].
No comments:
Post a Comment