• The Journal of the Acoustical Society of Korea
• /
• v.19 no.5
• /
• pp.20-26
• /
• 2000

In speaker verification system the HMM(hidden Markov model) parameter updating using small amount of data and the priori threshold decision are crucial factor for dealing with long-term variability in people voices. In the paper we present the speaker model updating technique which can be adaptable to the session-to-intra speaker variability and the priori threshold determining technique. The proposed technique decreases verification error rates which the session-to-session intra-speaker variability can bring by adapting new speech data to speaker model parameter through Baum Welch re-estimation. And in this study the proposed priori threshold determining technique is decided by a hybrid score measurement which combines the world model based technique and the cohen model based technique together. The results show that the proposed technique can lead a better performance and the difference of performance is small between the posteriori threshold decision based approach and the proposed priori threshold decision based approach.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.491-494
• /
• 2001

실제 화자 인증기를 설계함에 있어서 발생하는 가장큰 문제는, 적은 Enrollment data로 화자 모델이 만들어 지므로 화자 인증기의 성능이 시간이 지남에 따라 굉장히 줄어들게 되는 것과, 미리 훈련된 데이터 만으로 Threshold를 설정함에 따라 차후 실제 사용 시에 발생하는 변이를 고려하지 못하여 역시 성능 저하의 문제를 발생시킨다는 것이다. 위의 문제를 해결하기 위해 이 논문은 화자 모델을 구성하는데 있어 MAP 방법을 적용하고, threshold를 Resetting하는 방법을 적용했다. 본 논문에서 제안한 방법으로 HTER값이 23%정도 줄어듦을 보여준다.

• ETRI Journal
• /
• v.32 no.2
• /
• pp.333-335
• /
• 2010

This letter proposes a non-coherent blind time-of-arrival (TOA) estimation scheme for impulse radio ultra-wideband systems. The TOA estimation is performed in two consecutive phases: the Rayleigh-quotient theorem-based coarse-signal acquisition (CSA) and the iterative-threshold-test-based fine time estimation (FTE). The proposed scheme serves in a blind manner without demanding any a priori knowledge of the channel and the noise. Analysis and simulations demonstrate that the proposed scheme significantly increases the signal detection probability in CSA and ameliorates the TOA estimation accuracy in FTE.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1992.06a
• /
• pp.123-126
• /
• 1992

A new class of data transformation matri is introduced for estimation of angles of arrivals by the rank reduction of multiple wideband sources. The proposed unitary focusing matri minimizes the average of the squared norm of focusing error over the angles of interest without a priori knowledge of source locations. The merit that result as a consequence is a lower resolution threshold. These matrices can be applied to the case of the multigroup sources. Simulations and the comparison of statistical performance are compared with the algorithms (especially, spatial resampling method) which does not require the pre-estimation.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.24 no.3
• /
• pp.504-509
• /
• 1987

In this paper a characteristics of image which can be segmented based on the thresholding technique using a histogram was investigated employing 3 parameters: the variance of pixel value, the average mean difference between target and background and the target size. The threshold value for the histogram segmentation was determined by applying the hypothesis testing theory. The performance of the selected threshold was evaluated by computing a probability of error. Since a priori probability can be easily obtained from the histogram, it was found that the Bayes decision rule which theoretically guarantees the minimum probability of error works better than the minimax criterion rule.

• The Journal of the Acoustical Society of Korea
• /
• v.30 no.7
• /
• pp.383-389
• /
• 2011

In this paper, we propose a improved speech absence probability estimation algorithm by applying environmental noise classification for speech enhancement. The previous speech absence probability required to seek a priori probability of speech absence was derived by applying microphone input signal and the noise signal based on the estimated value of a posteriori SNR threshold. In this paper, the proposed algorithm estimates the speech absence probability using noise classification algorithm which is based on Gaussian mixture model in order to apply the optimal parameter each noise types, unlike the conventional fixed threshold and smoothing parameter. Performance of the proposed enhancement algorithm is evaluated by ITU-T P.862 PESQ (perceptual evaluation of speech quality) and composite measure under various noise environments. It is verified that the proposed algorithm yields better results compared to the conventional speech absence probability estimation algorithm.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.1
• /
• pp.16-24
• /
• 2007

Voice activity detectors (VADs) are important in wireless communication and speech signal processing. In the conventional VAD methods. an expression for the likelihood ratio test (LRT) based on statistical models is derived. Then, speech or noise is decided by comparing the value of the expression with a threshold. We propose a new method with the modified decision rule based on the Gaussian distribution and the uniformly most power (UMP) test. This method requires the distribution of the absolute value of the incoming speech signal. Then we can obtain the final decision through the relation between the Rayleigh distributions. This VAD method can detect speech without a priori signal-to-noise ratio (SNR) which is required in the conventional VAD algorithms. Additionally, in the various VAD performance tests, the proposed VAD method is shown to be more effective than the traditional scheme.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.47 no.8
• /
• pp.42-46
• /
• 2010

Energy detection based spectrum sensing compares the energy of a received signal from a primary user with a detection threshold and decides whether it is active or not in the frequency band of interest. Here the detection threshold depends on not only a target false alarm probability but also the level of the noise energy in the band. So, if the noise energy changes, the detection threshold must be adjusted accordingly to maintain the given false alarm probability. Most previous works on energy detection for spectrum sensing are based on the assumption that noise energy is known a priori. In this paper, we present a new energy detection scheme updating its detection threshold under the assumption that the noise is white, and analyze its detection performance. Analytic results show that the proposed scheme can maintain a target false alarm rate without regard to the noise energy level and its spectrum sensing performance gets better as the time bandwidth product of the signal used to estimate the noise energy increases.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.48 no.11
• /
• pp.22-27
• /
• 2011

In spectrum sensing, an energy detector compares the energy of a received signal with a predetermined detection threshold and decides whether a primary user is active or not in a licensed frequency band. Here the detection threshold is related to the noise power level in the band. Most previous works on energy detection have assumed that the noise power is exactly known a priori. However, this assumption does not hold in practice since there may be some uncertainty about the noise power. So it is necessary to investigate its effects on the performance of energy detection for spectrum sensing. In this paper, we analyze the effects using the residue theorem for contour integral and present the associated numerical results.

• ETRI Journal
• /
• v.41 no.2
• /
• pp.176-183
• /
• 2019

Spectrum sensing plays an important role in spectrum sharing. Energy detection is generally used because it does not require a priori knowledge of primary user (PU) signals; however, it is sensitive to noise uncertainty. An order statistics (OS) detector provides inherent protection against nonhomogeneous background signals. However, no analysis has been conducted yet to apply OS detection to spectrum sensing in a wireless channel to solve noise uncertainty. In this paper, we propose a robust spectrum sensing scheme based on generalized order statistics (GOS) and analyze the exact false alarm and detection probabilities under noise uncertainty. From the equation of the exact false alarm probability, the threshold value is calculated to maintain a constant false alarm rate. The detection probability is obtained from the calculated threshold under noise uncertainty. As a fusion rule for cooperative spectrum sensing, we adopt an OR rule, that is, a 1-out-of-N rule, and we call the proposed scheme GOS-OR. The analytical results show that the GOS-OR scheme can achieve optimum performance and maintain the desired false alarm rates if the coefficients of the GOS-OR detector can be correctly selected.