Discrimination Method of Synthetic Speech Using Pitch Frequency against Synthetic Speech Falsification.
Akio OgiharaHitoshi UnnoAkira ShiozakiPublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2005)
Keyphrases
- neural network
- fundamental frequency
- speaker recognition
- preprocessing
- experimental evaluation
- significant improvement
- computational cost
- real world
- linear prediction
- classification method
- high accuracy
- probabilistic model
- pairwise
- computational complexity
- objective function
- dynamic programming
- pattern recognition
- detection method
- speech corpus
- face recognition
- similarity measure
- dialogue system
- hybrid method
- error rate
- clustering method
- em algorithm
- classification accuracy
- evolutionary algorithm
- prior knowledge