SEARNN: Training RNNs with Global-Local Losses.
Rémi LeblondJean-Baptiste AlayracAnton OsokinSimon Lacoste-JulienPublished in: CoRR (2017)
Keyphrases
- feedforward neural networks
- echo state networks
- recurrent neural networks
- training examples
- training algorithm
- test set
- probabilistic model
- hidden markov models
- training set
- training process
- object detection
- training samples
- databases
- feed forward
- computer vision
- learning algorithm
- recurrent networks
- training program
- neural network