作者: Yang Gao , None
DOI: 10.1121/1.2040264
关键词: Physics 、 Acoustics 、 Signal transfer function 、 Multiplicative noise 、 Analog signal 、 Stochastic resonance 、 Noise floor 、 Phase noise 、 Noise (signal processing) 、 Frequency domain 、 Acoustics and Ultrasonics 、 Arts and Humanities (miscellaneous)
摘要: An input signal enters a noise suppression system in time domain and is converted to frequency domain. The then estimates ratio of the signal. Next, gain calculated based on estimated voicing parameter. parameter may be determined or ahead with respect time. In that event, fed back system, for example, by speech coder, calculate gain. After calculating gain, modifies using enhance quality. modified further from coding.