Michele Gubian

Learn More
Articulatory and acoustic reduction can manifest itself in the temporal and spectral domains. This study introduces a measure of spectral reduction, which is based on the speech decoding techniques commonly used in automatic speech recognizers. Using data for four frequent Dutch affixes from a large corpus of spontaneous face-to-face conversations, it(More)
Creating stimuli for perceptual experiments in intonation research involves manipulation of pitch contours extracted from spoken utterances. Difficulties arise when changes in the contour shape need to be applied globally and smoothly in the whole pitch curve. Moreover, it is hard to relate a gradual modification in some contour trait to its perceptual(More)
In this work we propose the use of Functional Data Analysis (FDA) as a powerful methodology to tackle problems where multiple continuous speech parameters have to be analyzed jointly. A production study on contrastive focus placement in Neapolitan Italian is used as illustration. Two features are analyzed, viz. f0 and relative speech rate, both expressed as(More)
In this paper we introduce Functional Data Analysis (FDA) as a tool for analyzing dynamic transitions in speech signals. FDA makes it possible to perform statistical analyses of sets of math­ ematical functions in the same way as classical multivariate analysis treats scalar measurement data. We illustrate the use of FDA with a reduction phenomenon(More)
In this paper, we face the problem of designing accurate decision-making modules in measurement systems that need to be implemented on resource-constrained platforms. We propose a methodology based on multiobjective optimization and genetic algorithms (GAs) for the analysis of support vector machine (SVM) solutions in the classification error-complexity(More)
Intonational research is often dependent upon hand-labeling by trained listeners, which can be prone to bias or error. We apply tools from Functional Data Analysis (FDA) to a set of fundamental frequency (F0) data to demonstrate how these tools can provide a less theory-dependent way of investigating F0 contours by allowing statistical analyses of whole(More)
Making accurate verbatim transcriptions is very time­ consuming and in the case of extemporaneous speech of native and non-native speakers the task is extremely difficult. While previous research focused on evaluating phonemic transcrip­ tions, the goal of our research is the automatic detection of tran­ scription errors on the orthographic level, which(More)
In this work we focus on the design of reduced-complexity sensor compensation modules based on learning-from-examples techniques. A Multi-Objective Optimization design framework is proposed, where system complexity and compensation uncertainty are considered as two conflicting costs to be jointly minimized. In addition, suitable statistical techniques are(More)
This study investigates the mechanism of tonal contraction when a disyllabic unit is merged into a monosyllable at fast speech rate in Taiwan Mandarin. Various degrees of contraction of bi-tonal sequences were elicited by manipulating speech rates. Functional Data Analysis was performed to compare trajectories of F0 and velocity in the contracted and(More)