Semi-supervised Acoustic Modelling for Five-lingual Code-switched ASR using Automatically-segmented Soap Opera Speech.
Nick WilkinsonAstik BiswasEmre YilmazFebe de WetEwald van der WesthuizenThomas NieslerPublished in: SLTU/CCURL@LREC (2020)
Keyphrases
- automatic speech recognition
- automatically segmented
- semi supervised
- speech sounds
- speech recognition
- acoustic features
- speech signal
- emotion recognition
- speech recognizers
- emotional speech
- acoustic models
- speech recognition systems
- word error rate
- semi supervised learning
- web services
- speech segments
- speaker independent
- spontaneous speech
- broadcast news
- hidden markov models
- prosodic features
- vocal tract
- speech synthesis
- speech corpus
- description language
- speaker verification
- source code
- noisy environments
- speech retrieval
- unlabeled data
- spoken words
- back end
- conversational speech
- labeled data
- active learning
- supervised learning
- formant frequencies
- acoustic signal
- language model
- pattern recognition
- speech recognizer
- semi supervised clustering
- audio visual
- multi view
- pairwise
- human computer interaction
- digital images
- decision trees