Noise

About: Noise is a research topic. Over the lifetime, 5111 publications have been published within this topic receiving 69407 citations. The topic is also known as: Мопсы танцуют под радио бандитов из сталкера 10 часов.

Speech audiometry in quiet with the Oldenburg Sentence Test for Children.

Abstract: Objective: To assess speech perception in children, speech audiometric sentence tests are generally better suited than single word tests because of their steeper discrimination function and thus higher sensitivity. A disadvantage of older German single word speech audiometric tests for children is their inapplicability in quiet and in noise. Moreover, their discrimination functions are shallower than those of optimized sentence tests, particularly in noise. The Oldenburg sentence test for children (Oldenburger Kinder-Satztest; OlKiSa) has already been shown to test reliably the speech perception in noise in normal-hearing children. Testing hearing-impaired children in noise, however, may be difficult. Therefore, quality criteria and norms for testing in quiet are also needed. Study sample: The OlKiSa in quiet was validated with 224 normal-hearing children between ages 4 to 10 years. Results: The discrimination functions are steeper (6.4 to 10.7 %/dB) than those of the commonly used German single w...

Clean Audio for TV broadcast: an object-based approach for hearing impaired viewers

Abstract: A significant proportion of any population suffer from some form of hearing loss, and age demographics mean that this proportion is increasing. A prominent effect of hearing loss is an increased difficulty to pick out speech signals in the presence of background noise or music, which can be particularly problematic for broadcast audio content. Consequently a common subjectofcomplaints tobroadcastersisthatthedialogueismaskedbybackground noise/music or that speech intelligibility is poor. This paper presents these issues, reviews previously proposed solutions, and presents an object-based approach that can be implemented within MPEG-H that can give users control of their audio mix and hence facilitate enhancements to be made for hearing impaired viewers.

Noise-like signals and their detection by correlation

Abstract: Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering, 1952.

Adaptive two-band spectral subtraction with multi-window spectral estimation

Abstract: An improved spectral subtraction algorithm for enhancing speech corrupted by additive wideband noise is described. The artifactual noise introduced by spectral subtraction that is perceived as musical noise is 7 dB less than that introduced by the classical spectral subtraction algorithm of Berouti et al. (1979). Speech is decomposed into voiced and unvoiced sections. Since voiced speech is primarily stochastic at high frequencies, the voiced speech is high-pass filtered to extract its stochastic component. The cut-off frequency is estimated adaptively. Multi-window spectral estimation is used to estimate the spectrum of stochastically voiced and unvoiced speech, thereby reducing the spectral variance. A low-pass filter is used to extract the deterministic component of voiced speech. Its spectrum is estimated with a single window. Spectral subtraction is performed with the classical algorithm using the estimated spectra. Informal listening tests confirm that the new algorithm creates significantly less musical noise than the classical algorithm.

Can't touch this! representations of the African American female body in urban rap videos

Abstract: (1997). Can't touch this! representations of the African American female body in urban rap videos. Popular Music and Society: Vol. 21, Cartographies of Sound, Noise, and Music at Century' End, pp. 107-116.