THE EVALUATION OF THE FEMININE LEVEL OF SPEECH

Wang, Yiqiu (2023) THE EVALUATION OF THE FEMININE LEVEL OF SPEECH. Master thesis, Voice Technology (VT).

Preview

PDF MSc_Voice_Tech YiqiuWang (final).pdf Download (271kB) | Preview

Abstract

The assessment of femininity and masculinity levels within speech constitutes a significant attribute perceptible to human listeners. This study delves into a novel methodology for quantifying these gender-related aspects of speech by leveraging the Deep Speaker model in tandem with Support Vector Machine (SVM) techniques. The outcome of this evaluation manifests as the "feminine level" of the speech. Through a comparative analysis between the predicted feminine level derived from the model and the outcomes of a listening test, which captures human perceptions of femininity and masculinity of speech, a discernible positive linear correlation emerges. Consequently, this study concludes that the proposed technique effectively predicts feminine levels, offering promising implications for applications centered around speech assessment and synthesis.

Item Type:	Thesis (Master)
Name supervisor:	Nayak, S.
Date Deposited:	12 Sep 2023 11:09
Last Modified:	12 Sep 2023 11:09
URI:	https://campus-fryslan.studenttheses.ub.rug.nl/id/eprint/369

Actions (login required)

View Item