On the Capability of Neural Networks to Generalize to Unseen Category-Pose Combinations.
Spandan MadanTimothy HenryJamell DozierHelen HoNishchal BhandariTomotake SasakiFrédo DurandHanspeter PfisterXavier BoixPublished in: CoRR (2020)
Keyphrases
- neural network
- fine grained categorization
- pose estimation
- pattern recognition
- artificial neural networks
- neural network model
- learning capabilities
- feed forward
- fuzzy logic
- genetic algorithm
- d objects
- rule extraction
- data sets
- pose variations
- object categories
- back propagation
- self organizing maps
- multilayer perceptron
- recurrent neural networks
- fault diagnosis
- position and orientation
- activation function
- training examples
- text classification
- previously unseen
- cellular neural networks
- adaptive resonance theory
- training set