Exploring nuisance attribute projection and score normalization for GLDS-SVM based automatic mispronunciation detection method

In the task of mispronunciation detection, the cross-speaker degradation and some other confusing nuisances are the challenging problems demanding prompt solution. In this paper, we will attempt to remove the non-pronunciation variations in the GLDS-SVM expansion space by using nuisance attribute projection strategy, in order to increase the separating capacity between different phoneme instances. Moreover, different kinds of score normalization methods with softmax, posterior probability vector (PPV), Z-norm and T-norm are comparatively discussed. The experiments on three kinds of speech corpora demonstrate the effectiveness of the above methods, and the performance mprovement is not very significant, but sustainable.i

• Type: Journal
• Year: 2011
• Where: ICASSP
• Authors: Hongyan Li, Shen Huang, Shijin Wang, Jiaen Liang, Bo Xu
• Keywords: ICASSP 2011, Nuisance Attribute Projection, Posterior Probability Vector, Score Normalization Methods, Signal Processing