Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: time-frequency masking

Sortuj według:

Ogranicz wyniki do:

A Data Field method for speech enhancement incorporating Binary Time-Frequency Masking

Huang J., Zhang Y., Zhang X., Zhu T.

Przegląd Elektrotechniczny

2011

R. 87, nr 7

225-229

A data field approach coupled with binary time-frequency masking is presented for the speech enhancement problem. In this proposed approach, data field method is employed to model the time and frequency dependencies of speech. This formulation has proved to be very helpful in enhancing speech quality by exploiting the correlation of speech both in time and in frequency. The experimental results demonstrate that the proposed algorithm offers improved signal to noise ratio and less spectral distortion.

Do poprawy jakości dźwięku mowy zastosowano metodę pola danych (Data field) połączoną z binarnym maskowanie czasowoczęstotliwościowym. Pozwoliło to znacząco poprawić jakość dźwięku przez wykorzystanie korelacji czasowej i częstotliwościowej. Uzyskano poprawę stosunku sygnału do szumu i zmniejszenie poziomu zniekształceń.

Relaxing the WDO Assumption in Blind Extraction of Speakers from Speech Mixtures

Kasprzak W., Ding N., Hamada N.

Journal of Telecommunications and Information Technology

2010

nr 4

50-58

The time-frequency masking approach in blind speech extraction consists of two main steps: feature clustering in a space spanned over delay-time and attenuation rate, and spectrogram masking in order to reconstruct the sources. Usually a binary mask is generated under the strong W-disjoint orthogonal (WDO) assumption (disjoint orthogonal representations in the frequency domain). In practice, this assumption is most often violated leading to weak quality of reconstructed sources. In this paper we propose the WDO to be relaxed by allowing some frequency bins to be shared by both sources. As we detect instantaneous fundamental frequencies the mask creation is supported by exploring a harmonic structure of speech. The proposed method is proved to be effective and reliable in experiments with both simulated and real acquired mixtures.