rotation invariant neural-network based face detection
DESCRIPTION
Rotation Invariant Neural-Network Based Face Detection. Overview. Multiple Neural Networks Router Networks Detector Networks. Overview of how the algorithm works. Input and output of the router network. Rotation Network: Outputs are generated as weighted vectors. - PowerPoint PPT PresentationTRANSCRIPT
Rotation Invariant Neural-Network
Based Face Detection
Rotation Invariant Neural-Network
Based Face Detection
OverviewOverview
Multiple Neural Networks Router Networks Detector Networks
Multiple Neural Networks Router Networks Detector Networks
Overview of how the algorithm works
Overview of how the algorithm works
Input and output of the router networkInput and output of the router network
Rotation Network:Outputs are generated as weighted
vectors
Rotation Network:Outputs are generated as weighted
vectors Average of the weighted vectors is interpreted
as an angle 1048 training images labeled by face, eyes, tip
of the nose, corners and centers of the mouth Each training face is rotated 15 times in a
circle
Average of the weighted vectors is interpreted as an angle
1048 training images labeled by face, eyes, tip of the nose, corners and centers of the mouth
Each training face is rotated 15 times in a circle
Rotation Neural Net Description
Rotation Neural Net Description
400 layers on the input layer (20X20)
Hidden layer of 15 units, output layer of 36 units.
Hyperbolic tangent activation function
Standard error back propigation
400 layers on the input layer (20X20)
Hidden layer of 15 units, output layer of 36 units.
Hyperbolic tangent activation function
Standard error back propigation
Detector NetworkDetector Network
Identical to the routing network. Trained by positive (contains faces) and negative images (does not contain faces).
Weights are initially random for the first training iteration.
Training on non-face images, add false positives to the non-image
Identical to the routing network. Trained by positive (contains faces) and negative images (does not contain faces).
Weights are initially random for the first training iteration.
Training on non-face images, add false positives to the non-image
Adding False Positives to the training set as negative
images
Adding False Positives to the training set as negative
images
Arbitration SchemeArbitration Scheme
Detection of Different Faces at different angles in the same image
Detections are placed in 4 dimensional space - x,y,angle, pyramid level, quantized in 10 degree increments.
Two independently trained networks are ANDed to improve the success rate.
Detection of Different Faces at different angles in the same image
Detections are placed in 4 dimensional space - x,y,angle, pyramid level, quantized in 10 degree increments.
Two independently trained networks are ANDed to improve the success rate.
Empirical Results:Empirical Results:
130 images, 511 faces 130 images, 511 faces
Sample Images
ConclusionsConclusions
Represents ways of integration multiple neural nets
Speed of implementation Face Detection VS Facial Recognition
Represents ways of integration multiple neural nets
Speed of implementation Face Detection VS Facial Recognition