Multilingual Jointly Trained Acoustic and Written Word Embeddings.
Yushi HuShane SettleKaren LivescuPublished in: INTERSPEECH (2020)
Keyphrases
- speech recognition systems
- language specific
- prosodic features
- co occurrence
- parallel corpus
- language independent
- vector space
- digital libraries
- n gram
- dimensionality reduction
- machine translation system
- training set
- source localization
- machine translation
- speech recognition
- low dimensional
- cross lingual
- word sense disambiguation
- keywords
- word segmentation
- word recognition
- bilingual dictionaries
- text to speech
- acoustic features
- writing style
- multilayer perceptron