SparseSpeech: Unsupervised Acoustic Unit Discovery with Memory-Augmented Sequence Autoencoders.
Benjamin MildeChris BiemannPublished in: INTERSPEECH (2019)
Keyphrases
- denoising
- unsupervised learning
- knowledge discovery
- semi supervised
- data driven
- data sets
- discovery process
- memory usage
- computational power
- pattern discovery
- main memory
- scientific discovery
- source localization
- data mining
- memory size
- underwater acoustic
- pseudorandom
- limited memory
- memory space
- supervised learning
- active learning
- training data
- neural network