Towards Visually Grounded Sub-word Speech Unit Discovery.
David HarwathJames R. GlassPublished in: ICASSP (2019)
Keyphrases
- speech recognition
- speech recognizer
- speech recognition systems
- prosodic features
- text input
- lexical features
- spontaneous speech
- co occurrence
- spoken document retrieval
- english text
- spoken documents
- recognition errors
- automatic speech recognition
- knowledge discovery
- text to speech
- speech recognizers
- spoken language
- word recognition
- word error rate
- broadcast news
- automatic transcription
- data mining
- n gram
- speech signal
- scientific discovery
- dialogue system
- audio visual
- keyword spotting
- discovery process
- noisy environments
- speech synthesis
- speaker recognition
- word sense disambiguation
- conversational speech
- word pairs
- related words
- spoken term detection
- grapheme to phoneme conversion