Comparative analysis between HMM and GMM-UBM methods:: search for α-optimal child voiceovers using the VTLN technique
A technical comparison of HMM and GMM-UBM methods for child voice processing, investigating VTLN-based optimization techniques.
About This Book
This work examines two statistical modeling approaches for voice processing tasks.
It focuses on adapting models for child speech using vocal tract length normalization.
The study compares performance between hidden Markov models and Gaussian mixture model-universal background models.
Emphasis is placed on identifying optimal parameters for improved accuracy in child voice applications.
Reviews
No reviews yet. Be the first to review this book!