A study of the automatic discrimination of emotion in three different time windows of speech is presented. A large corpus of acted emotional speech in Finnish was collected in five emotional states: neutral, sadness, joy, anger and tenderness. Automatic emotion discrimination tests were performed in speaker-independent scenarios using kNN classifiers with sequential forwards-backwards floating ...

The pure-tone audiogram gives a limited information about the r esidual hearing of hearing-impaired subjects. This i s especially the case when the hearing loss is severe or profound. In these cases often wide variations a r e obtained in results on speech-discrimination tests for the same degree of hearing loss. In some cases it i s likely that the puretone audiogram only shows tactile sensati...

This report investigates the use of expected attainable discrimination (EAD) as a measure to select discrete valued features in two-class prediction problems. In essence, EAD tells us the performance we could expect to achieve with a simple histogram probability density model of a given dataset. For discrete valued features, this kind of density model is bias-free but can have large variance. G...

A robust and effective voice activity detection (VAD) algorithm is proposed for improving speech recognition performance in noisy environments. The approach is based on filtering the input channel to avoid high energy noisy components and then the determination of the speech/non-speech bispectra by means of third order autocumulants. This algorithm differs from many others in the way the decisi...

We investigate the discrimination of phrase final pitch contours within a continuum from statement to question in English. Previous work in German [14] and Dutch [13] has raised questions about the relationship between discrimination sensitivity and category structure within this continuum. To clarify the relationship between linguistic category and simple auditory discrimination, we employ bot...

The problem of speech/noise discrimination has become increasingly important as the automatic speech recognition (ASR) system is applied in the real world. Robustness and simplicity are two challenges to the speech/noise discrimination method for an embedded system. The energy-based feature is the most suitable and applicable feature for speech/noise discrimination for embedded ASR system becau...

Automatic Speech Recognition (ASR) is widely used in many applications and tools. Smartphones, video games, cars are a few examples where people use ASR routinely often daily. A less commonly used, but potentially very important arena for using ASR, the health domain. For some people, impact on life could be enormous. The goal of this work to develop an easy-to-use, non-invasive, inexpensive sp...

BACKGROUND speech perception of elderly, in silence and in the presence of background noise. AIM to characterize and compare the hearing abilities of elderly in a monaural speech perception test, in the presence and absence of background noise (Speech Discrimination Test - SDT and Speech Perception in Noise - SPIN). METHOD Participants of this study were 55 subjects of both genders, 60 year...