Design and Implementation of Speech Enhancement System Based on Microphone Array
|School||Xi'an University of Electronic Science and Technology|
|Course||Communication and Information System|
|Keywords||Speech enhancement Microphone array Time delay estimation Beamforming|
In most cases, speech is inevitably corrupted acoustically by ambient noise. The noise degrades the speech SNR and its intelligibility, which affects the performance of most speech processing system significantly and results in a system abnormity at the worst time. The speech enhancement technique is an effective way to suppress the interference. Due to the information from single microphone system is limited to temporal/spectral, the performance is not satisfactory. To solve this problem, spatial and temporal/spectral information can be jointly exploited by using microphone.Speech enhancement system based on microphone array is one of the most popular research topics in recent years. The speech enhancement system based on microphone array involves in two key techniques: time delay estimation and beamforming. The work in this paper focuses on these two techniques. In the aspect of time delay estimation, the performance of phase transform weighted generalized cross-correlation method is analyzed. The performance analysis is given under -20dB ~ 10dB SNR. The microphone array database is used to analyze performance in actual circumstances. The precision of algorithm in digital signal processing is analyzed; it is proved that the algorithm is practicality. In the aspect of beamforming, delay and sum beamformer is improved to achieve constant beam width in voice band. The performance of suppressing low-frequency interference is improved. Finally, this paper presents a viable microphone array speech enhancement system design. The implementation of the software and hardware of the system is dealt with in detail.