Patch-based Fake Fingerprint Detection Using a Fully Convolutional Neural Network with a Small Number of Parameters and an Optimal Threshold.
Eunsoo ParkXuenan CuiWeonjin KimJinsong LiuHakil KimPublished in: CoRR (2018)
Keyphrases
- small number
- convolutional neural network
- threshold values
- detection method
- optimal parameters
- maximum likelihood
- dynamic programming
- object detection
- detection algorithm
- detection rate
- false positives
- face detection
- machine learning
- adaptive threshold
- parameter settings
- automatic detection
- parameter estimation
- linear combination
- optimal solution
- hough transform
- closed form
- worst case
- parameter values
- pairwise
- false alarms
- window size
- object recognition
- scaling factors
- image segmentation
- neural network
- sigmoid function