دانلود رایگان مقاله انگلیسی تقویت گفتار با یک فیلتر وینر تطبیقی به همراه ترجمه فارسی
عنوان فارسی مقاله: | تقویت گفتار با یک فیلتر وینر تطبیقی |
عنوان انگلیسی مقاله: | Speech enhancement with an adaptive Wiener filter |
رشته های مرتبط: | مهندسی کامپیوتر، مهندسی برق، هوش مصنوعی، مهندسی الکترونیک، مخابرات میدان |
فرمت مقالات رایگان | مقالات انگلیسی و ترجمه های فارسی رایگان با فرمت PDF میباشند |
کیفیت ترجمه | کیفیت ترجمه این مقاله خوب میباشد |
توضیحات | ترجمه این مقاله به صورت خلاصه انجام شده است. |
نشریه | اسپرینگر – Springer |
کد محصول | f163 |
مقاله انگلیسی رایگان |
دانلود رایگان مقاله انگلیسی |
ترجمه فارسی رایگان |
دانلود رایگان ترجمه مقاله |
جستجوی ترجمه مقالات | جستجوی ترجمه مقالات |
بخشی از ترجمه فارسی مقاله: –فیلتر وینرتطبیقی |
بخشی از مقاله انگلیسی: 5 Adaptive Wiener filtering This proposed adaptive Wiener filter benefits from the varying local statistics of the speech signal. A block diagram of the adaptive Wiener filtering method is illustrated in Fig. 1. In the filtering process, the estimated local mean mx and local variance σ2 x of the signal x(n) are exploited 6 Simulation results For the evaluation purpose, we have used a speech signal for the sentence “We were away year ago” for a male and for a female. We used speech quality metrics such as the SNR, segmental SNR (SNRseg), Log-Likelihood Ratio (LLR) and Spectral Distortion (SD). We first began with the male signal and added AWGN to it with SNRs of −5 and 5 dB. The results of all enhancement methods explained above on the male speech signal for SNR of 5 dB are shown in Figs. 2 to 8.For the colored noise case, we simulated colored noise by lowpass filtering of the AWGN prior to adding it to the signal. We also tested all speech enhancement methods on the male and the female signals in the presence of colored noise. The results of the tests for the male signal at SNR equal 5 dB are shown in Figs. 9 to 14. Figure 15 shows the output SNR versus the input SNR for all methods on the male signal. Figure 16 shows the SNRseg versus the input SNR for all methods on the male signal. Figure 17 shows the variation of the LLR versus the input SNR for all methods on the male signal. Figure 18 shows the variation of the SD versus the input SNR for all methods on the male signal. The case of the colored noise has also been studied in the comparison and its results are given in Figs 19 to 22. A similar study has been repeated on the female signal, and the results are tabulated in Tables 1 to 4.These results are all in favor of the proposed adaptive Wiener filtering method 7 Conclusion An adaptive Wiener filtering method for speech enhancement has been presented in this paper. This method depends on the adaptation of the filter impulse response from sample to sample based on the speech signal statistics. The results show that the proposed adaptive Wiener filtering method has the best performance as compared to all other speech enhancement methods mentioned in this paper at both low and high SNR values. The proposed filter succeeds in both the AWGN and the colored noise cases. This is attributed to the adaptive nature of the filter impulse response. This proposed adaptive Wiener filter has another advantage of being dependent only on the noisy signal as a single input. |