In this paper we present the multimodal data collected for developing a system able to influence the behavior of small groups in an informal and non goal-oriented conversation scenario. The prototype system looks like a table in a museum cafeteria and it is aimed at inducing the people sitting around to talk about their visit to the museum. To this aim, the system provides visual cues to foster...

This paper presents the behavioural mechanism of a statistical modelbased voice activity detector (VAD), featuring a likelihood ratio test for the activity decision. From investigation of the VAD, it is found that detection errors could occur frequently at speech offset regions because of the delay term in the decision-directed parameter estimator, employed for the estimation of an unknown para...

This paper describes some methodological issues to be considered when designing systems for automatic detection of voice pathology, in order to allow comparisons with previous or future experiments. The proposed methodology is built around Kay Elemetrics voice disorders database, which is the only one commercially available. Discussion about key points on this database is included. Any experime...

Separating technique for singing voice from music accompaniment is very useful in original sound type Karaoke instrument. We propose a real-time system to separate singing voice from music accompaniment for stereo recordings. Proposed algorithm consists of two stages. The first stage is a spectral change detector. The last stage is a selective vocal separation in frequency bins. Our system cons...

This study looks into the distribution of creaky voice in Mandarin in continuous speech. A creaky voice detector was used to automatically detect the appearance of creaky voice in a large-scale Mandarin corpus (Sinica COSPRO corpus). As the prosodic information has been annotated in the corpus, we were able to look at the distribution of creaky voice as a function of the interaction between ton...

In this work some modifications to the evaluation of PRMA protocols are presented. First, a more realistic voice model which takes in account the hangover of the speech detector is used, based on ITU-T P-84 recommendation. Second, the Front End Clipping (FEC) is calculated and taken as a limiting parameter instead of the dropping probability. Unlike this last, the FEC is a standard parameter to...

A lie detector which can reveal lie and deception in some automatic and perfectly reliable way is an old idea we have often met with in science fiction books and comic strips. This is all very well. It is when machines claimed to be lie detectors appear in the context of criminal investigations or security applications that we need to be concerned. In the present paper we will describe two type...

There is an increased risk for vocal and voice diseases due to the modern way of life. It is well known that most of the vocal and voice diseases cause changes in the acoustic voice signal. These diseases have to be diagnosed and treated during an early stage. Acoustic analysis is a non-invasive technique based on digital processing of speech signal. Acoustic analysis can be a useful tool to di...