Acoustic Word Embeddings for Untranscribed Target Languages with Continued Pretraining and Learned Pooling.
Ramon SanabriaOndrej KlejchHao TangSharon GoldwaterPublished in: CoRR (2023)
Keyphrases
- speech recognition systems
- statistical machine translation
- expressive power
- n gram
- language independent
- english text
- compound words
- grammar induction
- target language
- language specific
- vector space
- cross lingual
- co occurrence
- source localization
- previously learned
- manifold learning
- low dimensional
- word forms
- high dimensional
- prosodic features
- speech recognizers
- lexical features
- character n grams
- indian languages
- machine translation system
- word recognition
- word segmentation