An auto-encoder based approach to unsupervised learning of subword units.
Leonardo BadinoClaudia CanevariLuciano FadigaGiorgio MettaPublished in: ICASSP (2014)
Keyphrases
- unsupervised learning
- rate distortion
- bit rate
- supervised learning
- dimensionality reduction
- low complexity
- motion estimation
- semi supervised
- noisy channel
- machine learning
- distributed video coding
- model selection
- expectation maximization
- probabilistic model
- reinforcement learning
- computer vision
- high dimensional
- training data
- deep architectures