Combining Manual and Automatic Prosodic Annotation for Expressive Speech Synthesis.
Sandrine BrognauxThomas FrançoisMarco SaerensPublished in: LREC (2016)
Keyphrases
- speech synthesis
- manual annotation
- speech recognition
- prosodic features
- text to speech
- labor intensive
- semi automatic
- automatic annotation
- fully automatic
- error prone
- vocal tract
- automatic indexing
- semi automatically
- automatic analysis
- image annotation
- automatic image annotation
- semantic annotation
- case study
- fully automated
- manually annotated
- real time
- speech corpus
- multi modal
- hidden markov models
- image retrieval
- automatic speech recognition
- language model
- active learning
- hand crafted
- image processing
- learning algorithm
- genetic algorithm
- information retrieval