Home HomeABSTRACT
In many situations of digital speech communication (e.g. hands-free telephony or electronic hearing aids) the speech signal picked up by the microphone is disturbed by acoustic background noise. Therefore, adaptive filtering techniques which aim at the reduction of the disturbing noise are subject of current research activities (e.g. [1]). Although some of the already known adaptive techniques -- especially concepts with two or more microphones -- allow a significant reduction of the noise, most of the adaptive strategies result, particularly at low SNR, in a speech signal with an unnatural character due to time-variant distortions, and the occurrence of musical noise. An alternative approach, which does not affect the speech signal by time-variant distortions, is the application of a microphone array with a fixed directivity pattern aligned to the speaker's position, resulting in a suppression of spatially distributed noise sources. In this contribution it is shown that due to the proposed optimization even with only two microphones a reduction of diffuse noise sound up to 6 dB can be achieved.
A0072.pdf
TOPABSTRACT
One of the key technologies for natural man-machine interface is hands-free speech recognition. The performance of hands-free distant-talking speech recognition will be seriously degraded by noise and reverberation in real environments. A microphone array is applied to solve the problem. When applying a microphone array to speech recognition, parameters such as number of microphone elements and their spacing interval affect the performance. In order to optimize these parameters, a measure which reflects recognition performance is needed. In this paper, we investigate a measure of a microphone array design for speech recognition through experiments using various kinds of a microphone array design.
A0146.pdf
TOPABSTRACT
This paper proposes a front-end method for enhancing the target signal by subtracting estimated noise from a noisy signal using paired microphones, assuming that the noise is unevenly distributed with regard to time, frequency, and the direction. Although the Griffiths-Jim type adaptive beamformer has been proposed using the same concept, this method has some drawbacks. For example, sudden noises cannot be reduced because the convergence speed of the adaptive filter is slow; also the signal is distorted in a reverberated environment. The proposed method, however, can over-come the above drawbacks by formulating noises using arrival time differences between paired microphones and by estimating noises analytically using the directions of the noises. The simulated results show that the method with one paired microphone can increase signal-to-noise ratios (SNR) by 10 ~ 20 dB in simulations and can reduce log-spectrum distances by about 5 dB in real noisy environments.
A0208.pdf
TOPABSTRACT
This paper addresses the problem of enhancing a speech signal corrupted by interfering signals. A new noise reduction algorithm based on a logarithmic microphone array and the multiresolution wavelet transform is described. The proposed processing is applied in the time-spectral domain with respect to the logarithmic subband decomposition of the spectrum of each microphone signal. The advantage of the proposed method is that both the sub-array based beamforming operation and the postfiltering are performed in the same transform domain without adding FFT processing. Computer simulation results show that our approach is effective for noise reduction. The technique can be used in hands-free voice communication applications operating in an adverse environment. In particular, it can be applied to improve speech signal pick-up for voice communication terminal.
A0342.pdf
TOPABSTRACT
This paper proposes a new robust adaptive beamformer suitable for a two-channel microphone array. The proposed beamformer is a combination of two Griffith-Jim generalized sidelobe cancellers (GSC) in which the look direction of each GSC is determined by the other GSC's directional response. Moreover, to reduce the degradation of interference suppression performance caused by spatial aliasing, a new arrangement of directional microphones is proposed. The arrangement is simple and effectively reduces the degradation. Simulations demonstrate the effectiveness of the proposed two-channel two-beamformer microphone array.
A0358.pdf
TOPABSTRACT
In this work hands-free continuous speech recognition based on microphone arrays is investigated. A set of experiments was carried out using arrays having different numbers of omnidirectional microphones as well as different configurations. Both real and simulated array signals, generated by means of the image method, were used. An enhanced input to a recognizer based on Hidden Markov Models was obtained by a time delay compensation module providing a beamformed signal. HMM adaptation was used to realign the recognizer acoustic modeling to the given acoustic condition.
A0690.pdf
TOP