• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2017.11a
• /
• pp.202-203
• /
• 2017

비음수 행렬 인수분해(Non-negative Matrix Factorization, NMF)의 신호분리 성능을 개선하기 위해 희소조건을 인가한 방법이 희소 비음수 행렬 인수분해 알고리즘(Sparse NMF, SNMF)이다. 기존의 SNMF 알고리즘은 개별 음원의 희소성을 고려하지 않고 임의로 결정한 희소 조건을 사용한다. 본 논문에서는 음원의 특성에 따른 희소성을 추정하고 이를 SNMF 학습알고리즘에 적용하는 새로운 신호분리 기법을 제안한다. 혼합 신호에서의 잡음제거 실험을 통해, 제안한 방법이 기존의 NMF와 SNMF에 비해 성능이 더 우수함을 보였다.

• Journal of Information Technology Applications and Management
• /
• v.26 no.6
• /
• pp.89-101
• /
• 2019

Many researchers have been focused on designing beauty product recommendation system for a long time because of increased need of customers for personalized and customized recommendation in beauty product domain. In addition, as the application of the deep neural network technique becomes active recently, various collaborative filtering techniques based on the deep neural network have been introduced. In this context, this study proposes a deep neural network model suitable for beauty product recommendation by applying Neural Collaborative Filtering and Generalized Matrix Factorization (NCF + GMF) to beauty product recommendation. This study also provides an implementation of web API system to commercialize the proposed recommendation model. The overall performance of the NCF + GMF model was the best when the beauty product recommendation problem was defined as the estimation rating score problem and the binary classification problem. The NCF + GMF model showed also high performance in the top N recommendation.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.2
• /
• pp.123-129
• /
• 2017

This paper proposes a new nonnegative matrix factorization (NMF) based direction-of-arrival (DOA) estimation method for multiple sound sources using a dual microphone array. First of all, sound signals coming from the dual microphone array are segmented into consecutive analysis frames, and a steered-response power phase transform (SRP-PHAT) beamformer is applied to each frame so that stereo signals of each frame are represented in a time-direction domain. The time-direction outputs of SRP-PHAT are stored for a pre-defined number of frames, which is referred to as a time-direction block. Next, In order to estimate DOAs robust to noise, each time-direction block is normalized along the time by using a block subtraction technique. After that, an unsupervised NMF method is applied to the normalized time-direction block in order to cluster the directions of each sound source in a multiple sound source environments. In particular, the activation and basis matrices are used to estimate the number of sound sources and their DOAs, respectively. The DOA estimation performance of the proposed method is evaluated by measuring a mean absolute error (MAE) and the standard deviation of errors between the oracle and estimated DOAs under a three source condition, where the sources are located in [$-35{\circ}$, 5m], [$12{\circ}$, 4m], and [$38{\circ}$, 4.m] from the dual microphone array. It is shown from the experiment that the proposed method could relatively reduce MAE by 56.83%, compared to a conventional SRP-PHAT based DOA estimation method.

• Proceedings of the Korea Air Pollution Research Association Conference
• /
• 2000.04a
• /
• pp.291-293
• /
• 2000

대기 연구자들은 대기오염의 일반적인 현황과 대기오염 유발의 근본 원인 파악, 저감 대책 등에 대한 연구를 활발히 수행하고 있다 하지만 이러한 연구들 중에서 대기오염의 근본 원인을 파악하기 위한 오염원 (Source) 추정 연구는 국내외적으로 매우 미진하다. 대기질의 평가와 예측은 분산모델과 수용방법론을 이용하는데, 분산모델에 내재되어 있는 한계성과 제약점 때문에 수용체에서 오염물질의 특성을 분석한 후, 오염원의 기여도를 평가하는 수용방법론이 지속적으로 개발되고 있다. (중략)

• Journal of Korea Multimedia Society
• /
• v.10 no.3
• /
• pp.303-315
• /
• 2007

This paper proposes a 3D structure recovery and registration method that uses four or more common points. For each frame of a given video, a partial structure is recovered using tracked points. The 3D coordinates, camera positions and camera directions are computed at once by our improved scaled orthographic factorization method. The partially recovered point sets are parts of a whole model. A registration of point sets makes the complete shape. The recovered subsets are integrated by transforming each coordinate system of the local point subset into a common basis coordinate system. The process of shape recovery and integration is performed uniformly and linearly without any nonlinear iterative process and without loss of accuracy. The execution time for the integration is significantly reduced relative to the conventional ICP method. Due to the fast recovery and registration framework, our shape recovery scheme is applicable to various interactive video applications. The processing time per frame is under 0.01 seconds in most cases and the integration error is under 0.1mm on average.

• Journal of Korean Society for Atmospheric Environment
• /
• v.24 no.1
• /
• pp.100-107
• /
• 2008

The objective of this study was to quantitatively estimate PM-10 source contribution in Gumi City, Korea. Ambient PM-10 samples were collected by a high volume air sampler, which operated for 84 different days with a 24-h sampling basis, from June 14,2001 though May 19, 2003. The filter samples were analyzed for determining 13 inorganic elements, 3 anions, and a total carbon. The study has intensively applied a receptor model, the PMF (Positive Matrix Factorization) model. The results from PMF modeling indicated that a total of seven sources were independently identified and each source was contributed to the ambient Gumi City from secondary sulfate (34%), motor vehicle (26%), soil relation (5%), field burning (3%), industrial relation (3%), secondary nitrate (22%), and incinration (7%) in terms of PM-10 mass, respectively.

• ETRI Journal
• /
• v.38 no.3
• /
• pp.487-493
• /
• 2016

Two new methods are proposed for an unsupervised adaptation of a language model (LM) with a single sentence for automatic transcription tasks. At the training phase, training documents are clustered by a method known as Latent Dirichlet allocation (LDA), and then a domain-specific LM is trained for each cluster. At the test phase, an adapted LM is presented as a linear mixture of the now trained domain-specific LMs. Unlike previous adaptation methods, the proposed methods fully utilize a trained LDA model for the estimation of weight values, which are then to be assigned to the now trained domain-specific LMs; therefore, the clustering and weight-estimation algorithms of the trained LDA model are reliable. For the continuous speech recognition benchmark tests, the proposed methods outperform other unsupervised LM adaptation methods based on latent semantic analysis, non-negative matrix factorization, and LDA with n-gram counting.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.18 no.12
• /
• pp.1408-1418
• /
• 2007

In this paper, we analyze the performance of pilot symbol assisted channel estimation and equalization schemes for OFDM systems over frequency-selective time-varying channels and propose methods to improve the system performance. In the least square(LS) and linear minimum mean square error(MMSE) channel estimation, time domain windowing is introduced for banding the frequency domain channel matrix. The linear MMSE and decision feedback equalization schemes are employed with the pilot symbols for channel estimation taken into account in the equalization process. To reduce computational complexity, the band LU matrix factorization algorithm is introduced in solving the linear systems involved in the equalization, and the performances are compared with the known previous results by computer simulations. When time domain windowing is employed in the decision feedback equalization, the matrix related with the decision feedback process is shown to be unhanded and the resultant performance degradation is analyzed.

• Journal of Advanced Navigation Technology
• /
• v.19 no.6
• /
• pp.544-551
• /
• 2015

In this paper, the design and implementation results of the optical flow estimator (OFE) for moving object detection (MOD) in advanced driver assistance system (ADAS). In the proposed design, Brox's algorithm with global optimization is considered, which shows the high performance in the vehicle environment. In addition, Cholesky factorization is applied to solve Euler-Lagrange equation in Brox's algorithm. Also, shift register bank is incorporated to reduce memory access rate. The proposed optical flow estimator was designed with Verilog-HDL, and FPGA board was used for the real-time verification. Implementation results show that the proposed optical flow estimator includes the logic slices of 40.4K, 155 DSP48s, and block memory of 11,290Kbits.

• Journal of Korean Society for Atmospheric Environment
• /
• v.22 no.1
• /
• pp.85-97
• /
• 2006

The PAMS data collected at four sites in Seoul metropolitan area in 2004 were analyzed using the positive matrix factorization (PMF) technique, in order to identify the possible sources and estimate their contributions to ambient VOCs. Ten sources were then resolved at Jeongdong, Bulgwang, Yangpyeong, and Seokmo, including vehicle exhaust, LPG vehicle, petroleum evaporation, coating, solvent, asphalt, LNG, Industry & heating, open burning, and biogenic source. The PMF analysis results showed that vehicle exhaust commonly contributed the largest portion of the predicted total VOCs mass concentration, more than $30\%$ at four sites. The contribution of other resolved sources were significantly different according to the characteristics of site location. In the case of Jeongdong and bulgwang located in urban area, various anthropogenic sources such as coating, solvent, asphalt, residual LPG, and petroleum evaporation contributed about $40\%$ of total VOCs mass. On the other hand, at yangpyeong and Seokmo located in rural and remote area, the portion of these anthropogenic sources was reduced to less than $30\%$ and the contribution of natural sources including open burning and biogenic source clearly observed. These results were considerably corresponding to the emission inventory investigated in this region.