Presence Of Non-stationary Noise Research Articles

BCI-Walls: A robust methodology to predict if conscious EEG changes can be detected in the presence of artefacts.

Brain computer interfaces (BCI) depend on reliable realtime detection of conscious EEG changes for example to control a video game. However, scalp recordings are contaminated with non-stationary noise, such as facial muscle activity and eye movements. This interferes with the detection process making it potentially unreliable or even impossible. We have developed a new methodology which provides a hard and measurable criterion if conscious EEG changes can be detected in the presence of non-stationary noise by requiring the signal-to-noise ratio of a scalp recording to be greater than the SNR-wall which in turn is based on the highest and lowest noise variances of the recording. As an instructional example, we have recorded signals from the central electrode Cz during eight different activities causing non-stationary noise such as playing a video game or reading out loud. The results show that facial muscle activity and eye-movements have a strong impact on the detectability of EEG and that minimising both eye-movement artefacts and muscle noise is essential to be able to detect conscious EEG changes.

A comparative study of neural network methods for first break detection using seismic refraction data over a detrital iron ore deposit

Accurate basement delineation is crucial for mineral deposits where formation is associated with a paleo-basement topography, such as in detrital iron ore deposits. In such environments seismic surveys provide useful data to assist the mapping of the basement interface. A key stage of the process is identification of the first arrival of elastic waves at a recording station, or so-called ‘first break’. Although a variety of automated first break detection algorithms were proposed previously for use in different seismological environments, first break detection in hard rock environments remains challenging due to low amplitude arrivals and the presence of non-stationary noise within the seismic traces. This study explores the use of three types of supervised neural networks for first break detection for a seismic refraction dataset of 15,000 traces, over a detrital iron ore deposit. A fully connected neural network (NN), a Convolutional Neural Network (CNN), and a Long-Short Term Memory recurrent neural network (LSTM) are evaluated and their predicted first breaks are compared to manually picked first breaks and predictions from the widely used Coppens’ method. The results showed that the CNN and LSTM were particularly effective for the first break detection, achieving mean square errors of 0.051 ms (with a statistical bias of 2.2 ms) and 0.058 (with a statistical bias of 21.2 ms) respectively against manual interpretation, compared to –35 ms for Coppens’ method and x ms for the NN −122 ms.

Investigation of a New Weighted Averaging Method to Improve SNR of Electrocochleography Recordings.

The aim of this study was to investigate methods to improve signal-to-noise ratio (SNR) of extratympanic electrocochleography (ET-ECOG); a low SNR electrophysiological measurement technique. The current standard for ET-ECOG involves acquiring and uniform averaging ∼1000 evoked responses to reveal the signal of interest. Weighted averaging is commonly employed to enhance SNR of repetitive signals in the presence of a nonstationary noise, yet its efficacy in ET-ECOG has not been explored to date, which was the focus of this study. Conventional techniques used to compute signal statistics required for weighted averaging were found to be ineffective for ET-ECOG due to low SNR; therefore, a modified correlation coefficient-based approach was derived to quantify the "signal" component. Several variants of weighted averaging schemes were implemented and evaluated on 54 ECOG recordings obtained from seven healthy volunteers. The best weighted averaging scheme provided a 17% (p < 0.05) SNR increase [signal amplitude to standard deviation (STD) of the noise ratio] compared to uniform averaging, and further improved to 22% (p < 0.05) when variance of the noise was incorporated as a cost factor. The implemented weighted averaging schemes were robust and effective for variants of ET-ECOG recording protocols investigated. Weighted averaging improved SNR of low amplitude ET-ECOG recordings in the presence of nonstationary noise. SNR improvements for ECOG have significant benefits in clinical applications; the variability associated with biofeatures extracted can be reduced, and may lead to shorter recordings. Methods described in this study can easily be incorporated in other low SNR repetitive electrophysiological measurement techniques.

A Multistream Feature Framework Based on Bandpass Modulation Filtering for Robust Speech Recognition.

There is strong neurophysiological evidence suggesting that processing of speech signals in the brain happens along parallel paths which encode complementary information in the signal. These parallel streams are organized around a duality of slow vs. fast: Coarse signal dynamics appear to be processed separately from rapidly changing modulations both in the spectral and temporal dimensions. We adapt such duality in a multistream framework for robust speaker-independent phoneme recognition. The scheme presented here centers around a multi-path bandpass modulation analysis of speech sounds with each stream covering an entire range of temporal and spectral modulations. By performing bandpass operations along the spectral and temporal dimensions, the proposed scheme avoids the classic feature explosion problem of previous multistream approaches while maintaining the advantage of parallelism and localized feature analysis. The proposed architecture results in substantial improvements over standard and state-of-the-art feature schemes for phoneme recognition, particularly in presence of nonstationary noise, reverberation and channel distortions.

Biomimetic multi-resolution analysis for robust speaker recognition.

Humans exhibit a remarkable ability to reliably classify sound sources in the environment even in presence of high levels of noise. In contrast, most engineering systems suffer a drastic drop in performance when speech signals are corrupted with channel or background distortions. Our brains are equipped with elaborate machinery for speech analysis and feature extraction, which hold great lessons for improving the performance of automatic speech processing systems under adverse conditions. The work presented here explores a biologically-motivated multi-resolution speaker information representation obtained by performing an intricate yet computationally-efficient analysis of the information-rich spectro-temporal attributes of the speech signal. We evaluate the proposed features in a speaker verification task performed on NIST SRE 2010 data. The biomimetic approach yields significant robustness in presence of non-stationary noise and reverberation, offering a new framework for deriving reliable features for speaker recognition and speech processing.

Noise robust voice activity detection based on periodic to aperiodic component ratio

This paper proposes a noise robust voice activity detection (VAD) technique called PARADE (PAR based Activity DEtection) that employs the periodic component to aperiodic component ratio (PAR). Conventional noise robust features for VAD are still sensitive to non-stationary noise, which yields variations in the signal-to-noise ratio, and sometimes requires a priori noise power estimations, although the characteristics of environmental noise change dynamically in the real world. To overcome this problem, we adopt the PAR, which is insensitive to both stationary and non-stationary noise, as an acoustic feature for VAD. By considering both periodic and aperiodic components simultaneously in the PAR, we can mitigate the effect of the non-stationarity of noise. PARADE first estimates the fundamental frequencies of the dominant periodic components of the observed signals, decomposes the power of the observed signals into the powers of its periodic and aperiodic components by taking account of the power of the aperiodic components at the frequencies where the periodic components exist, and calculates the PAR based on the decomposed powers. Then it detects the presence of target speech signals by estimating the voice activity likelihood defined in relation to the PAR. Comparisons of the VAD performance for noisy speech data confirmed that PARADE outperforms the conventional VAD algorithms even in the presence of non-stationary noise. In addition, PARADE is applied to a front-end processing technique for automatic speech recognition (ASR) that employs a robust feature extraction method called SPADE (Subband based Periodicity and Aperiodicity DEcomposition) as an application of PARADE. Comparisons of the ASR performance for noisy speech show that the SPADE front-end combined with PARADE achieves significantly higher word accuracies than those achieved by MFCC (Mel-frequency Cepstral Coefficient) based feature extraction, which is widely used for conventional ASR systems, the SPADE front-end without PARADE, and other standard noise robust front-end processing techniques (ETSI ES 202 050 and ETSI ES 202 212). This result confirmed that PARADE can improve the performance of front-end processing for ASR.

Robust endpoint detection algorithm based on the adaptive band-partitioning spectral entropy in adverse environments

In speech processing, endpoint detection in noisy environments is difficult, especially in the presence of nonstationary noise. Robust endpoint detection is one of the most important areas of speech processing. Generally, the feature parameters used for endpoint detection are highly sensitive to the environment. Endpoint detection is severely degraded at low signal-to-noise ratios (SNRs) since those feature parameters cannot adequately describe the characteristics of a speech signal. As a result, this study seeks the banded structure on speech spectrogram to distinguish a speech from a nonspeech, especially in adverse environments. First, this study proposes a feature parameter, called band-partitioning spectral entropy (BSE), which exploits the use of the banded structure on speech spectrogram. A refined adaptive band selection (RABS) method is extended from the adaptive band selection method proposed by Wu et al., which adaptively selects useful bands not corrupted by noise. The successful RABS method is strongly depended on an on-line detection with minimal processing delay. In this paper, the RABS method is combined with the BSE parameter. Finally, a novel robust feature parameter, adaptive band-partitioning spectral entropy (ABSE), is presented to successfully detect endpoints in adverse environments. Experimental results indicate that the ABSE parameter is very effective under various noise conditions with several SNRs. Furthermore, the proposed algorithm outperforms other approaches and is reliable in a real car.

Noise adaptive speech recognition based on sequential noise parameter estimation

In this paper, a noise adaptive speech recognition approach is proposed for recognizing speech which is corrupted by additive non-stationary background noise. The approach sequentially estimates noise parameters, through which a non-linear parametric function adapts mean vectors of acoustic models. In the estimation process, posterior probability of state sequence given observation sequence and the previously estimated noise parameter sequence is approximated by the normalized joint likelihood of active partial paths and observation sequence given the previously estimated noise parameter sequence. The Viterbi process provides the normalized joint-likelihood. The acoustic models are not required to be trained from clean speech and they can be trained from noisy speech. The approach can be applied to perform continuous speech recognition in presence of non-stationary noise. Experiments conducted on speech contaminated by simulated and real non-stationary noise show that when acoustic models are trained from clean speech, the noise adaptive speech recognition system provides improvements in word accuracy as compared to the normal noise compensation system (which assumes the noise to be stationary) in slowly time-varying noise. When the acoustic models are trained from noisy speech, the noise adaptive speech recognition system is found to be helpful to get improved performance in slowly time-varying noise over a system employing multi-conditional training.

Method of measuring signal parameters in the presence of nonstationary noise

A method, based on the maximum optimum maximum likelihood criterion, of measuring the parameters of signals containing quasi-deterministic and random noise is investigated.

Optimal detection using bilinear time-frequency and time-scale representations

Bilinear time-frequency representations (TFRs) and time-scale representations (TSRs) are potentially very useful for detecting a nonstationary signal in the presence of nonstationary noise or interference. As quadratic signal representations, they are promising for situations in which the optimal detector is a quadratic function of the observations. All existing time-frequency formulations of quadratic detection either implement classical optimal detectors equivalently in the time-frequency domain, without fully exploiting the structure of the TFR, or attempt to exploit the nonstationary structure of the signal in an ad hoc manner. We identify several important nonstationary composite hypothesis testing scenarios for which TFR/TSR-based detectors provide a natural framework; that is, in which TFR/TSR-based detectors are both optimal and exploit the many degrees of freedom available in the TFR/TSR. We also derive explicit expressions for the corresponding optimal TFR/TSR kernels. As practical examples, we show that the proposed TFR/TSR detectors are directly applicable to many important radar/sonar detection problems. Finally, we also derive optimal TFR/TSR-based detectors which exploit only partial information available about the nonstationary structure of the signal.

Optimal sensor location in the presence of nonstationary noise

Observability conditions for distributed parameter systems with nonstationary state excitation noise are derived. Because these conditions are violated in most cases of interest in process control an optimal projection approach and a sensor location criterion are developed to minimize the error caused by the lack of observability. The value of these concepts was demonstrated in real time experiments on a system with two spatial dimensions, the heating of a cylindrical ingot. It was found that through the nonstationary noise model and the optimal sensor location significantly improved state estimates can be obtained when persistent or slowly varying unmeasured disturbances affect the system. The real-time requirements for the filter were very modest and an application in the process industries where these disturbances always present a serious problem, seems entirely feasible.