Minimum generation error training for HMM-based prediction of articulatory movements.
Tian-Yi ZhaoZhen-Hua LingMing LeiLi-Rong DaiQingfeng LiuPublished in: ISCSLP (2010)
Keyphrases
- prediction error
- prediction accuracy
- hidden markov models
- prediction model
- training set
- training process
- supervised learning
- error rate
- training phase
- multi layer perceptron
- radial basis function network
- speech recognition
- error bounds
- motor skills
- training examples
- support vector
- relative error
- vocal tract
- square error
- general loss functions