A brain-computer interface (BCI) is a communication system that allows to control computer or any other device thanks the brain activity. The BCI described in this paper based on P300 speller paradigm introduced by Farwell and Donchin . An unsupervised algorithm proposed enhance evoked potentials estimating spatial filters; raw EEG signals are then projected into estimated signal subspace. Data recorded three subjects were used evaluate method. results, which presented using Bayesian linear...

In this paper, we present an extended nonlinear Bayesian filtering framework for extracting electrocardiograms (ECGs) from a single channel as encountered in the fetal ECG extraction abdominal sensor. The recorded signals are modeled summation of several ECGs. Each them is described by dynamic model, previously presented generation highly realistic synthetic ECG. Consequently, each has corresponding term model and can thus be efficiently discriminated even if waves overlap time. parameter...

A brain–computer interface (BCI) is a specific type of human–computer that enables direct communication between human and computer through decoding brain activity. As such, event-related potentials like the P300 can be obtained with an oddball paradigm whose targets are selected by user. This paper deals methods to reduce needed set EEG sensors in speller application. reduced number yields more comfort for user, decreases installation time duration, may substantially financial cost BCI setup...

The separation of speech signals measured at multiple microphones in noisy and reverberant environments using only the audio modality has limitations because there is generally insufficient information to fully discriminate different sound sources. Humans mitigate this problem by exploiting visual modality, which insensitive background noise can provide contextual about scene. This advantage inspired creation new field audiovisual (AV) source that targets alongside microphone measurements a...

Smart homes have been an active area of research, however despite considerable investment, they are not yet a reality for end-users. Moreover, there still accessibility challenges the elderly or disabled, two main potential targets home automation. In this exploratory study we design control mechanism smart based on Brain Computer Interfaces (BCI) and apply it in "Domus" platform order to evaluate interest users about BCIs at home. We enable lighting, TV set, coffee machine shutters...

Looking at the speaker's face can be useful to better hear a speech signal in noisy environment and extract it from competing sources before identification. This suggests that visual signals of (movements visible articulators) could used enhancement or extraction systems. In this paper, we present novel algorithm plugging audiovisual coherence signals, estimated by statistical tools, on audio blind source separation (BSS) techniques. is applied difficult realistic case convolutive mixtures....

A cross-sectional study was conducted on a random sample of 1,200 health care professionals in Marseille, France, order to assess the prevalences depression and burnout, compare these two entities. Depression assessed by Center for Epidemiologic Studies-Depression scale (CES-D), burnout Maslach Burnout Inventory (MBI). is syndrome emotional exhaustion, depersonalization towards patients, reduced sense personal accomplishment. Some psychiatrists consider be clinical form depression. The were...

We present a new approach to the voice activity detection (VAD) problem for speech signals embedded in non-stationary noise. The method is based on automatic lipreading: objective detect or non-activity by exploiting coherence between acoustic signal and speaker's lip movements. From comprehensive analysis of shape parameters during non-speech events, we show that single appropriate visual parameter, defined characterize movements, can be used sections more precisely, silence sections....

In this paper, we study the distribution of log-modulus a Gaussian complex random variable. circular case, it is Log-Rayleigh (LR) variable, whose probability function (pdf) depends on only one parameter. noncircular pdf more complicated, although show that can be adequately modeled by an LR pdf, for which optimal fitting parameter derived. These results used in any application using discrete Fourier transform coefficients, e.g., speech/audio signals, and suggest mixture kernels preferable...

This paper deals with coupled tensor factorization. A relaxed criterion derived from the advanced matrix-tensor factorization (ACMTF) proposed by Acar et al. is described. The ACMTF (RACMTF) based on weaker assumptions that are thus more often satisfied when dealing actual data. Numerical simulations show benefit of using jointly two data sets underlying factors highly correlated, especially if one modality less noisy than other one. method finally applied Gaze&EEG to estimate ocular...

This paper presents a quantitative and comprehensive study of the lip movements given speaker in different speech/nonspeech contexts, with particular focus on silences (i.e., when no sound is produced by speaker). The aim to characterize relationship between "lip activity" "speech then use visual speech information as voice activity detector (VAD). To this aim, an original audiovisual corpus was recorded two speakers involved face-to-face spontaneous dialog, although being separate rooms....

The Eye Fixation Related Potential (EFRP) estimation is the average of EEG signals across epochs at ocular fixation onset. Its main limitation overlapping issue. Inter Intervals (IFI) - typically around 300 ms in case unrestricted eye movement- depend on participants’ oculomotor patterns, and can be shorter than latency components evoked potential. If duration an epoch longer IFI value, more one occur, some between adjacent neural responses ensues. classical does not take into account either...

This letter introduces a new technique for phonocardiogram (PCG) signal denoising based on nonnegative matrix factorization (NMF) of its spectrogram and adaptive contour representation computation (ACRC) short-time Fourier transform (STFT). More precisely, NMFs PCG synchronous electrocardiogram spectrograms are first used to filter out high-energy noises from PCG. Then, ACRC is performed low-pass filtered version the STFT resulting identify relevant time-frequency components that...

In this paper we present two novel methods for visual voice activity detection (V-VAD) which exploit the bimodality of speech (i.e. coherence between speaker's lips and resulting speech). The first method uses appearance parameters a lips, obtained from an active model (AAM). An HMM then dynamically models change in over time. second retinal filter on region to extract required parameter. A corpus single speaker is applied each turn, where used classify as or non speech. efficiency evaluated...