• The Journal of the Acoustical Society of Korea
• /
• v.25 no.2
• /
• pp.78-87
• /
• 2006

In this paper. we propose an effective mask estimation scheme for missing-feature reconstruction in order to achieve robust speech recognition under unknown noise environments. In the previous work. colored noise is used for training the mask classifer, which is generated from the entire frequency Partitioned signals. However it gives a limited performance under the restricted number of training database. To reflect the spectral events of more various background noise and improve the performance simultaneously. a new Bayesian classifier for mask estimation is proposed, which works independent of other frequency bands. In the proposed method, we employ the colored noise which is obtained by combining colored noises generated from each frequency band in order to reflect more various noise environments and mitigate the 'sparse' database problem. Combined with the cluster-based missing-feature reconstruction. the performance of the proposed method is evaluated on a task of noisy speech recognition. The results show that the proposed method has improved performance compared to the Previous method under white noise. car noise and background music conditions.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.5
• /
• pp.38-43
• /
• 1996

This paper compares the recognition performances of feature vectors known to be robust to the environmental noise. And, the speech subtraction technique is combined with the noise robust feature to get more performance enhancement. The experiments using SMC(Short time Modified Coherence) analysis, root cepstral analysis, LDA(Linear Discriminant Analysis), PLP(Perceptual Linear Prediction), RASTA(RelAtive SpecTrAl) processing are carried out. An isolated word recognition system is composed using semi-continuous HMM. Noisy environment experiments usign two types of noises:exhibition hall, computer room are carried out at 0, 10, 20dB SNRs. The experimental result shows that SMC and root based mel cepstrum(root_mel cepstrum) show 9.86% and 12.68% recognition enhancement at 10dB in compare to the LPCC(Linear Prediction Cepstral Coefficient). And when combined with spectral subtraction, mel cepstrum and root_mel cepstrum show 16.7% and 8.4% enhanced recognition rate of 94.91% and 94.28% at 10dB.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 1996.04a
• /
• pp.263-270
• /
• 1996

Noise control of a plate structure with multiple disk shaped piezoelectric actuators is studied. The plate is excited by an acoustic pressure field produced by a noise source located below the plate. Finite element modeling is used for the plate structure that supports a combination of three dimensional solid, flat shell and transition elements. The objective function, in the optimization procedure, is to minimize the sound energy radiated onto a hemispherical surface of given radius and the design parameters are the locations and sizes of the piezoelectric actuators as well as the amplitudes of the voltages applied to them. Automatic mesh generation is addressed as part of the modeling procedure. Numerical results for both resonance and off resonance frequencies show remarkable noise reduction and the optimal locations of the actuators are found to be close to the edges of the plate structure. The optimized result is robust such that when the acoustic pressure pattern is changed, reduction of radiated sound is still maintained. The robustness of an optimally designed structure is also tested by changing the frequency of the noise source using only the actuator voltages as design parameters.

• Journal of Internet Computing and Services
• /
• v.8 no.4
• /
• pp.1-9
• /
• 2007

Decision fusion in wireless sensor networks under non-Gaussian noise channels is studied. To consider the tail behavior noise distributions, we use a exponentially-tailed distribution as a wide class of noise distributions. Based on a canonical parallel fusion model with fading and noise channels, the likelihood ratio(LR) based fusion rule is considered as an optimal fusion rule under Neyman-Pearson criterion. With both high and low signal-to-noise ratio (SNR) approximation to the optimal rule, we obtain several suboptimal fusion rules. and we propose a simple fusion rule that provides robust detection performance with a minimum prior information, Performance evaluation for several fusion rules is peformed through simulation. Simulation results show the robustness of the Proposed simple fusion rule.

• Journal of Digital Contents Society
• /
• v.18 no.8
• /
• pp.1603-1613
• /
• 2017

This paper proposes a method for effectively segmenting the hand region using a widely popular RGB color-based webcam. This performs the empirical preprocessing method four times to remove the noise. First, we use Gaussian smoothing to remove the overall image noise. Next, the RGB image is converted into the HSV and the YCbCr color model, and global fixed binarization is performed based on the statistical value for each color model, and the noise is removed by the bitwise-OR operation. Then, RDP and flood fill algorithms are used to perform contour approximation and inner area fill operations to remove noise. Finally, ROI (hand region) is selected by eliminating noise through morphological operation and determining a threshold value proportional to the image size. This study focuses on the noise reduction and can be used as a base technology of gesture recognition application.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.35 no.2
• /
• pp.175-181
• /
• 2011

We propose a novel algorithm for measuring the radius of tubes. This proposed algorithm is capable of effectively removing added noise and measuring the radius of tubes within allowable precision. The noise is removed by using a candidate true center that minimizes the standard deviation with respect to the radius. Further, the disconnection in data points resulting from noise removal is solved by using a connection algorithm. The final step of the process is repeated until the value of the standard deviation decreases to a small predefined value. Experiments were performed using circle geometries with added noise and a real tube with complex noise and that is used in the braking units of automobiles. It was concluded that the measurement carried out using the algorithm was accurate within 1.4%, even with 15% added noise.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.36 no.5
• /
• pp.423-432
• /
• 2023

With the increasing demand for mobile devices featuring multi-touch operation, extensive research is being conducted on touch screen panel (TSP) Readout ICs (ROICs) that should possess low power consumption, compact chip size, and immunity to external noise. Therefore, this paper discusses capacitive touch sensors and their readout circuits, and it introduces research trends in various circuit designs that are robust against external noise sources. The recent state-of-the-art TSP ROICs have primarily focused on minimizing the impact of parasitic capacitance (Cp) caused by thin panel thickness. The large Cp can be effectively compensated using an area-efficient current compensator and Current Conveyor (CC), while a display noise reduction scheme utilizing a noise-antenna (NA) electrode significantly improves the signal-to-noise ratio (SNR). Based on these achievements, it is expected that future TSP ROICs will be capable of stable operation with thinner and flexible Touch Screen Panels (TSPs).

• Convergence Security Journal
• /
• v.24 no.2
• /
• pp.201-207
• /
• 2024

Voice is a critical element of human communication, and the development of speech recognition models is one of the significant achievements in artificial intelligence, which has recently been applied in various aspects of human life. The application of speech recognition models in the military field is also inevitable. However, before artificial intelligence models can be applied in the military, it is necessary to research their vulnerabilities. In this study, we evaluates the military applicability of the multilingual speech recognition model "Whisper" by examining its vulnerabilities to battlefield noise, white noise, and adversarial attacks. In experiments involving battlefield noise, Whisper showed significant performance degradation with an average Character Error Rate (CER) of 72.4%, indicating difficulties in military applications. In experiments with white noise, Whisper was robust to low-intensity noise but showed performance degradation under high-intensity noise. Adversarial attack experiments revealed vulnerabilities at specific epsilon values. Therefore, the Whisper model requires improvements through fine-tuning, adversarial training, and other methods.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.9
• /
• pp.95-104
• /
• 2009

This paper proposes the noise robust algorithm to detect the starting point of music. Detection of starting point of music is necessary to solve computational-waste problem and retrieval-comparison problem with inconsistent input data in music content based retrieval system. In particular, such detection is even more necessary in time sequential retrieval method that compares data in the sequential order of time in contents based music retrieval system. Whereas it has the long point that the retrieval is fast since it executes simple comparison in the order of time, time sequential retrieval method has the short point that data starting time to be compared should be the same. However, digitalized music cannot guarantee the equity of starting time by bit rate conversion. Therefore, this paper ensured that recognition rate shall not decrease even while executing high speed retrieval by applying time sequential retrieval method through detection of music starting point in the pre-processing stage of retrieval. Starting point detection used minimum wave model that can detect effective sound, and for strength against noise, the noises existing in mute sound were swapped. The proposed algorithm was confirmed to produce about 38% more excellent performance than the results to which starting point detection was not applied, and was verified for the strength against noise.

• Journal of the Korean Data and Information Science Society
• /
• v.8 no.2
• /
• pp.163-171
• /
• 1997

Function approximation from a set of input-output pairs has numerous applications in scientific and engineering areas. Multilayer feedforward neural networks have been proposed as a good approximator of nonlinear function. The back propagation(BP) algorithm allows multilayer feedforward neural networks to learn input-output mappings from training samples. It iteratively adjusts the network parameters(weights) to minimize the sum of squared approximation errors using a gradient descent technique. However, the mapping acquired through the BP algorithm may be corrupt when errorneous training data we employed. When errorneous traning data are employed, the learned mapping can oscillate badly between data points. In this paper we propose a robust BP learning algorithm that is resistant to the errorneous data and is capable of rejecting gross errors during the approximation process, that is stable under small noise perturbation and robust against gross errors.