Xác định ngưỡng che cho thuật toán nâng cao chất lượng tiếng nói dựa vào thông tin sai lệch pha trong môi trường thực sử dụng hai microphone

• Tác giả: Quốc Cường Nguyễn
• Ngôn ngữ: vie
• Ký hiệu phân loại: 006.45 *Acoustical pattern recognition
• Thông tin xuất bản: Khoa học và Công nghệ các trường đại học kỹ thuật, 2013
• Mô tả vật lý: 28-33
• Bộ sưu tập: Metadata
• ID: 561086

Tóm tắt

Speech enhancement using one microphone has been researched and has obtained some good results. Most of speech enhancement approaches using one microphone are only suitable for stationary noise, however they are not appropriate with non-stationary noise. The speech enhancement approach using two microphones based on phase difference obtained in the frequency domain is a potential one for speech enhancement which is speech is masked by'other noise. This algorithm on previous systems that have been successful in separating signals based on differences in arrival time of signal components from two microphones in simulation environment. The paper aims to evaluate the performance of this approach in real environment and in other words, the authors proposed a manual masking threshold approach with prior knowledge position of target and position of noise.