摘要:
System and techniques for a microphone board for far field automatic speech recognition are described herein. The microphone board may include a first plurality of microphones disposed along a circumference of a circle on a surface and a second plurality of microphones disposed along a line on the surface. First connections to the first plurality of microphones may be grouped together and second connections to the second plurality of microphones are grouped together. The first connections and the second connections may be provided to an external entity of the surface via a connector.
摘要:
Embodiments of a system and method for adapting a phase difference-based noise reduction system are generally described herein. In some embodiments, spatial information associated with a first and second audio signal are determined, wherein the first and second audio signals including a target audio inside a beam and noise from outside the beam. A signal-to-noise ratio (SNR) associated with the audio signals is estimated. A mapping of phase differences to gain factors is adapted for determination of attenuation factors for attenuating frequency bins associated with noise outside the beam. Spectral subtraction is performed to remove estimated noise from the single-channel signal based on a weighting that affects frequencies associated with a target signal less. Frequency dependent attenuation factors are applied to attenuate frequency bins outside the beam to produce a target signal having noise reduced.
摘要:
System and techniques for automatic speech recognition pre-processing are described herein. First, a plurality of audio channels may be obtained. Then, reverberations mat be removed from the audio channels. The plurality of audio channels may be partitioned into beams after reverberations are removed. A partition corresponding to a beam in the beams may be selected based on a noise level. An audio signal may be filtered from the selected partition. The filtered audio signal may be provided to an external entity via an output interface of the pre-processing pipeline.
摘要:
System and techniques for automatic speech recognition de-reverberation are described herein. A portion of an audio stream may be obtained. here, the portion of the audio stream is a proper subset of the audio stream. A filter may be created by applying Generalized Weighted Prediction Error (GWPE) to the portion of the audio stream. The filter may be applied to the audio stream to remove reverberation. The filtered version of the audio stream may then be provided to an audio stream consumer.