Lithuanian Speech Recognition Using Purely Phonetic Deep Learning.
Laurynas PipirasRytis MaskeliunasRobertas DamaseviciusPublished in: Comput. (2019)
Keyphrases
- deep learning
- speech recognition
- spoken term detection
- speech sounds
- unsupervised learning
- speech signal
- unsupervised feature learning
- spoken document retrieval
- machine learning
- hidden markov models
- automatic speech recognition
- pattern recognition
- language model
- deep architectures
- broadcast news
- weakly supervised
- multiscale
- mental models
- restricted boltzmann machine
- data sets
- learning strategies
- input image
- higher order