Rapid speaker adaptation using regression-tree based spectral peak alignment
Shizhen Wang, Xiaodong Cui, Abeer A. Alwan · 2006
In this paper, regression-tree based spectral peak alignment is proposed for rapid speaker adaptation using the linearization of VTLN. Two different regression classes are investigated: phonetic classes (using combined knowledge and data-driven techniques) and mixture classes. Compared to MLLR and VTLN, improved performance can be obtained for both supervised and unsuper-vised adaptations on both medium vocabulary and connected dig-its recognition tasks. To further improve the performance, MLLR was integrated into this regression-tree based peak alignment. Ex-perimental results show that the performance improvements can be achieved even with limited adaptation data. Index Terms: speaker adaptation, peak alignment, regression tree 1.