Open Set Recognition Through Deep Neural Network Uncertainty: Does Out-of-Distribution Detection Require Generative Classifiers?
Martin MundtIuliia PliushchSagnik MajumderVisvanathan RameshPublished in: CoRR (2019)
Keyphrases
- neural network
- discriminative classifiers
- pattern recognition
- recognition rate
- object recognition
- recognition accuracy
- road signs
- scene text recognition
- handwritten digit recognition
- back propagation
- recognition algorithm
- multi layer perceptron
- traffic signs
- unsupervised learning
- generative model
- partial occlusion
- detection method
- detection algorithm
- feature extraction
- object detection
- probability distribution
- artificial neural networks
- test set
- action recognition
- anti noise
- discriminative learning
- nearest neighbour
- facial action units
- linear classifiers
- machine learning
- detection rate
- fuzzy logic
- multi class
- training set
- support vector
- feature selection
- genetic algorithm
- random variables
- machine learning algorithms
- face detection
- feature space
- training data
- boosted classifiers