Multilingual and crosslingual speech recognition using phonological-vector based phone embeddings.
Chengrui ZhuKeyu AnHuahuan ZhengZhijian OuPublished in: CoRR (2021)
Keyphrases
- vector space
- multi lingual
- text to speech synthesis
- emotional speech
- speech recognition
- automatically generated
- emotion recognition
- digital libraries
- acoustic models
- mobile phone
- automatic speech recognition
- manifold learning
- broadcast news
- text to speech
- correlation clustering
- euclidean space
- speaker identification
- spoken term detection
- language resources
- speech signal
- audio visual
- cross language
- high dimensional data
- low dimensional
- feature vectors
- spoken language
- speech synthesis
- cross lingual
- distance measure
- endpoint detection