Academia.eduAcademia.edu

Figure 2 – uploaded by IAES Journals

See full PDFdownloadDownload figure
Figure 2. Block diagram of proposed speech enhancement technique This paper describes a speech enhancement technique that is based on source separation and adaptive LMS to improve passenger voice commands derived from a signal containing varied interferences and different speech sounds. To begin, the microphone captures the necessary signals, which are then passed to the source separation, which separates the denoised signals and removes the permutation ambiguity. The frequency domain is used in the separation process. As a result, the convolutive mixes are converted to the frequency domain using a short-time Fourier transform (STFT). Finally, we use inverse short-time Fourier transform (ISTFT) to convert the unmixed signals to the time domain. The adaptive LMS receives the source separation outputs. Figure 2 depicts the specific procedure. 4.1. Input source mixture and preprocessing

Figure 2 Block diagram of proposed speech enhancement technique This paper describes a speech enhancement technique that is based on source separation and adaptive LMS to improve passenger voice commands derived from a signal containing varied interferences and different speech sounds. To begin, the microphone captures the necessary signals, which are then passed to the source separation, which separates the denoised signals and removes the permutation ambiguity. The frequency domain is used in the separation process. As a result, the convolutive mixes are converted to the frequency domain using a short-time Fourier transform (STFT). Finally, we use inverse short-time Fourier transform (ISTFT) to convert the unmixed signals to the time domain. The adaptive LMS receives the source separation outputs. Figure 2 depicts the specific procedure. 4.1. Input source mixture and preprocessing

Related Figures (8)

Positioning a microphone array inside a car needs to obey some important requisites related to response quality, installation costs, and user-friendliness. Several works dealing with in-car speect S t S t t eparation and enhancement adopt two main microphone dispositions: microphones spread throughout he whole car interior. Although presenting some interesting results, this type of disposition presents ome drawbacks that can make it difficult for its adoption in commercial systems. First, spreading he microphones throughout the car interior makes the receivers experiment with different levels of noise, produced by the horn sound, car engine sound, driver’s voice, the interruption from other passengers, and so on. This can lead to degradation in the speech signal response. Moreover, considering hat the objective is to enhance the speech of the passenger, microphones are placed on the headrests of t he front seats. The driver position, front-seat passenger position, and back-seat passenger position ar¢ S hown in Figure 1. Let X and Y denote the number of source signals and microphones, respectively. The signals fron he source are then referred to as (1).
Positioning a microphone array inside a car needs to obey some important requisites related to response quality, installation costs, and user-friendliness. Several works dealing with in-car speect S t S t t eparation and enhancement adopt two main microphone dispositions: microphones spread throughout he whole car interior. Although presenting some interesting results, this type of disposition presents ome drawbacks that can make it difficult for its adoption in commercial systems. First, spreading he microphones throughout the car interior makes the receivers experiment with different levels of noise, produced by the horn sound, car engine sound, driver’s voice, the interruption from other passengers, and so on. This can lead to degradation in the speech signal response. Moreover, considering hat the objective is to enhance the speech of the passenger, microphones are placed on the headrests of t he front seats. The driver position, front-seat passenger position, and back-seat passenger position ar¢ S hown in Figure 1. Let X and Y denote the number of source signals and microphones, respectively. The signals fron he source are then referred to as (1).
Figure 4. Operation of speech enhancement in car Table 1 lists the parameters that were used in the SSCA algorithm. Figure 5 shows the speech and different kinds of noise-mixed signals in the time domain and time-frequency domain, respectively. Human speech and ambient noise have a wide frequency distribution. They can provide a great deal of information about speech characteristics. Figures 6(a) and 6(b) illustrates the separation result of the desired signal in the time domain and spectrogram [21]. The noise signal of a moving car in the time domain and spectrogram is shown in Figure 6. The frequency of car sound is primarily distributed in 50-500 Hz with no apparent regularity across time. Figure 6 shows that the time-domain processed signals are significantly different from the sources, indicating interference signals can be found in the voice spectrum. Furthermore, the speech spectrogram contains noise-desirable signal information at 270 Hz that affects the speech enhancement, and the ambient signal has a high frequency which indicates the SSCA approach is dependent on the adaptive LMS algorithm. Figure 7(a) shows the separated interference speech signal and Figure 7(b) is the spectrogram correspondingly. The proposed technique improves the speech signal, as shown in Figures 8(a) and 8(b) shows the spectrogram. As the result is shown in Table 2 the proposed method improves speech more effectively. According to the performance results, an SSCA method gives a good SNR and perceptual evaluation of speech quality (PESQ) at each input source combination, when compared to the other methods. rable 1. Parameters used in source separation method
Figure 4. Operation of speech enhancement in car Table 1 lists the parameters that were used in the SSCA algorithm. Figure 5 shows the speech and different kinds of noise-mixed signals in the time domain and time-frequency domain, respectively. Human speech and ambient noise have a wide frequency distribution. They can provide a great deal of information about speech characteristics. Figures 6(a) and 6(b) illustrates the separation result of the desired signal in the time domain and spectrogram [21]. The noise signal of a moving car in the time domain and spectrogram is shown in Figure 6. The frequency of car sound is primarily distributed in 50-500 Hz with no apparent regularity across time. Figure 6 shows that the time-domain processed signals are significantly different from the sources, indicating interference signals can be found in the voice spectrum. Furthermore, the speech spectrogram contains noise-desirable signal information at 270 Hz that affects the speech enhancement, and the ambient signal has a high frequency which indicates the SSCA approach is dependent on the adaptive LMS algorithm. Figure 7(a) shows the separated interference speech signal and Figure 7(b) is the spectrogram correspondingly. The proposed technique improves the speech signal, as shown in Figures 8(a) and 8(b) shows the spectrogram. As the result is shown in Table 2 the proposed method improves speech more effectively. According to the performance results, an SSCA method gives a good SNR and perceptual evaluation of speech quality (PESQ) at each input source combination, when compared to the other methods. rable 1. Parameters used in source separation method
Figure 5. The (a) time domain and (b) spectrogram of input signal
Figure 5. The (a) time domain and (b) spectrogram of input signal
Figure 6. The (a) time domain and (b) spectrogram of separated desired signal
Figure 6. The (a) time domain and (b) spectrogram of separated desired signal
Figure 7. The (a) time domain and (b) spectrogram of separated interference signal
Figure 7. The (a) time domain and (b) spectrogram of separated interference signal
Figure 8. The (a) time domain and (b) spectrogram of enhanced desired signal
Figure 8. The (a) time domain and (b) spectrogram of enhanced desired signal
CONCLUSION In the car environment, a range of interference effects and varied speech signals provide a major difficulty for the operation of the hands-free system. In this paper, a proposed adaptive LMS and SSCA algorithm is used to process and enhance the speech signal from the desired direction. The input source direction is obtained using the TDOA between the microphones. The input signals are separated and reconstructed to get the desired interference signals. The separated signals are adaptively filtered and enhanced. The SSCA method is implemented and validated using real-time hardware with different combinations of input source mixtures, which has yielded the expected results. Real-time results show that the proposed technique can successfully separate various speech signals without the need for prior information in the car system. The efficiency of the system is then demonstrated using a two-microphone beamformer. The signal quality parameters of the input signal to the enhanced desired signal are SNR of 5.2 dB and PESQ of 2.3 achieved.
CONCLUSION In the car environment, a range of interference effects and varied speech signals provide a major difficulty for the operation of the hands-free system. In this paper, a proposed adaptive LMS and SSCA algorithm is used to process and enhance the speech signal from the desired direction. The input source direction is obtained using the TDOA between the microphones. The input signals are separated and reconstructed to get the desired interference signals. The separated signals are adaptively filtered and enhanced. The SSCA method is implemented and validated using real-time hardware with different combinations of input source mixtures, which has yielded the expected results. Real-time results show that the proposed technique can successfully separate various speech signals without the need for prior information in the car system. The efficiency of the system is then demonstrated using a two-microphone beamformer. The signal quality parameters of the input signal to the enhanced desired signal are SNR of 5.2 dB and PESQ of 2.3 achieved.
Related topics:
Speech enhancementMicrophone Arrays
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved