• 한국산학기술학회논문지
• /
• 제9권5호
• /
• pp.1249-1254
• /
• 2008

그래프 모델에서 가장 중요한 부분은 관찰 데이터가 주어진 상황에서 은닉 변수와 더불어 파라미터의 사후확률 분포의 계산이다. 이 논문에서는 가우시안 혼합 모델에 대한 변분 베이지안 방법의 구현과 변분 근사화 분포의 분해 유도를 제안한다. 이 방법은 정보 검색이나 데이터 시각화와 같은 데이터 분석 등에 적용이 가능하다.

• International Journal of Contents
• /
• 제9권1호
• /
• pp.1-5
• /
• 2013

Standard Gaussian Mixture Model (GMM) is a well-known method for image segmentation. However, one of its problems is that we consider the pixel as independent to each other, which can cause the segmentation results sensitive to noise. It explains why some of existing algorithms still cannot segment texts from the background clearly. Therefore, we present a new method in which we incorporate the spatial relationship between a pixel and its neighbors inside $3{\times}3$ windows to segment the text. Our approach works well with images containing texts, which has different sizes, shapes or colors in case of light changes or complex background. Experimental results demonstrate the robustness, accuracy and effectiveness of the proposed model in image segmentation compared to other methods.

• 대한전자공학회논문지SP
• /
• 제48권5호
• /
• pp.109-114
• /
• 2011

본 논문에서는 청각장애인을 위한 음향 데이터를 이용한 음향강화 알고리즘을 Gaussian Mixture Model (GMM)을 이용한 상황인지 시스템 기반으로 제안한다. 음향 신호 데이터에서 Mel-Frequency Cepstral Coefficients (MFCC) 특징벡터를 추출하여 GMM을 구성하고 이를 기반으로 상황인지 결과에 따라 위험음향일 경우 음향강화기술을 제안한다. 실험결과 제안된 상황인지 기반의 음향강화 알고리즘이 다양한 음향학적 환경에서 우수한 성능을 보인 것을 알 수 있었다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2015년도 추계학술발표대회
• /
• pp.1571-1572
• /
• 2015

본 논문에서는 GMM (Gaussian mixture model) 기반의 BS (background subtraction) 알고리즘을 이용한 이동 객체 검출기의 하드웨어 구조 설계 결과를 제시하였다. 설계된 이동객체 검출기는 1280 * 720 HD 해상도의 영상을 30 frames per second로 실시간 처리가 가능하다. 하드웨어 구현은 Verilog-HDL을 이용하였으며, FPGA 기반 구현 결과, 설계된 이동 객체 검출기는 582 Slice, 1,698 Slice LUT, 8 DSP48s, 1,769 Flip Flop, 691.2 KByte BRAM으로 구성되었음을 확인하였다.

• Journal of Information Processing Systems
• /
• 제9권4호
• /
• pp.621-632
• /
• 2013

The most important things for a forest fire detection system are the exact extraction of the smoke from image and being able to clearly distinguish the smoke from those with similar qualities, such as clouds and fog. This research presents an intelligent forest fire detection algorithm via image processing by using the Gaussian Mixture model (GMM), which can be applied to detect smoke at the earliest time possible in a forest. GMMs are usually addressed by making the model adaptive so that its parameters can track changing illuminations and by making the model more complex so that it can represent multimodal backgrounds more accurately for smoke plume segmentation in the forest. Also, in this paper, we suggest a way to classify the smoke plumes via a feature extraction using HSL(Hue, Saturation and Lightness or Luminanace) color space analysis.

• ETRI Journal
• /
• 제33권6호
• /
• pp.914-923
• /
• 2011

This paper concerns robust and reliable speaker model training for text-independent speaker verification. The baseline speaker modeling approach is the Gaussian mixture model (GMM). In text-independent speaker verification, the amount of speech data may be different for speakers. However, we still wish the modeling approach to perform equally well for all speakers. Besides, the modeling technique must be least vulnerable against unseen data. A traditional approach for GMM training is expectation maximization (EM) method, which is known for its overfitting problem and its weakness in handling insufficient training data. To tackle these problems, variational approximation is proposed. Variational approaches are known to be robust against overtraining and data insufficiency. We evaluated the proposed approach on two different databases, namely KING and TFarsdat. The experiments show that the proposed approach improves the performance on TFarsdat and KING databases by 0.56% and 4.81%, respectively. Also, the experiments show that the variationally optimized GMM is more robust against noise and the verification error rate in noisy environments for TFarsdat dataset decreases by 1.52%.

• Nuclear Engineering and Technology
• /
• 제56권2호
• /
• pp.715-727
• /
• 2024

During fast neutron imaging, besides the dark current noise and readout noise of the CCD camera, the main noise in fast neutron imaging comes from high-energy gamma rays generated by neutron nuclear reactions in and around the experimental setup. These high-energy gamma rays result in the presence of high-density gamma white spots (GWS) in the fast neutron image. Due to the microscopic quantum characteristics of the neutron beam itself and environmental scattering effects, fast neutron images typically exhibit a mixture of Gaussian noise. Existing denoising methods in neutron images are difficult to handle when dealing with a mixture of GWS and Gaussian noise. Herein we put forward a deep learning approach based on the Swin Transformer UNet (SUNet) model to remove high-density GWS-Gaussian mixture noise from fast neutron images. The improved denoising model utilizes a customized loss function for training, which combines perceptual loss and mean squared error loss to avoid grid-like artifacts caused by using a single perceptual loss. To address the high cost of acquiring real fast neutron images, this study introduces Monte Carlo method to simulate noise data with GWS characteristics by computing the interaction between gamma rays and sensors based on the principle of GWS generation. Ultimately, the experimental scenarios involving simulated neutron noise images and real fast neutron images demonstrate that the proposed method not only improves the quality and signal-to-noise ratio of fast neutron images but also preserves the details of the original images during denoising.

• 말소리와 음성과학
• /
• 제7권3호
• /
• pp.131-138
• /
• 2015

In this paper, we propose L1-norm regularization for state vector adaptation of subspace Gaussian mixture model (SGMM). When you design a speaker adaptation system with GMM-HMM acoustic model, MAP is the most typical technique to be considered. However, in MAP adaptation procedure, large number of parameters should be updated simultaneously. We can adopt sparse adaptation such as L1-norm regularization or sparse MAP to cope with that, but the performance of sparse adaptation is not good as MAP adaptation. However, SGMM does not suffer a lot from sparse adaptation as GMM-HMM because each Gaussian mean vector in SGMM is defined as a weighted sum of basis vectors, which is much robust to the fluctuation of parameters. Since there are only a few adaptation techniques appropriate for SGMM, our proposed method could be powerful especially when the number of adaptation data is limited. Experimental results show that error reduction rate of the proposed method is better than the result of MAP adaptation of SGMM, even with small adaptation data.

• Communications for Statistical Applications and Methods
• /
• 제9권3호
• /
• pp.715-723
• /
• 2002

This paper is to show a clustering application of a density estimation method that utilizes the Gaussian mixture model. We define "closeness measure" as a clustering criterion to see how close given two Gaussian components are. Closeness measure is defined as the ratio of log likelihood between two Gaussian components. According to simulations using artificial data, the clustering algorithm turned out to be very powerful in that it can correctly determine clusters in complex situations, and very flexible in that it can produce different sizes of clusters based on different threshold valuesold values

• ETRI Journal
• /
• 제32권5호
• /
• pp.795-800
• /
• 2010

Embedding a large vocabulary speech recognition system in mobile devices requires a reduced acoustic model obtained by eliminating redundant model parameters. In conventional optimization methods based on the minimum description length (MDL) criterion, a binary Gaussian tree is built at each state of a hidden Markov model by iteratively finding and merging similar mixture components. An optimal subset of the tree nodes is then selected to generate a downsized acoustic model. To obtain a better binary Gaussian tree by improving the process of finding the most similar Gaussian components, this paper proposes a new distance measure that exploits the difference in likelihood values for cases before and after two components are combined. The mixture weight of Gaussian components is also introduced in the component merging step. Experimental results show that the proposed method outperforms MDL-based optimization using either a Kullback-Leibler (KL) divergence or weighted KL divergence measure. The proposed method could also reduce the acoustic model size by 50% with less than a 1.5% increase in error rate compared to a baseline system.