• Journal of Multimedia Information System
• /
• v.8 no.1
• /
• pp.31-34
• /
• 2021

The classical HMM is defined by a parameter triple �� = (��, A, B), where each parameter represents a collection of probability distributions: initial state, state transition and output distributions in order. This paper proposes a new stationary parameter e = (e1, e2, …, eN) where N is the number of states and et = P(|xt = i, y) for describing how an input pattern y ends in state xt = i at time t followed by nothing. It is often said that all is well that ends well. We argue here that all should end well. The paper sets the framework for the theory and presents an efficient inference and training algorithms based on dynamic programming and expectation-maximization. The proposed model is applicable to analyzing any sequential data with two or more finite segmental patterns are concatenated, each forming a context to its neighbors. Experiments on online Hangul handwriting characters have proven the effect of the proposed augmentation in terms of highly intuitive segmentation as well as recognition performance and 13.2% error rate reduction.

• Journal of the Korean Society for Precision Engineering
• /
• v.12 no.4
• /
• pp.28-38
• /
• 1995

The flank wear at the minor cutting edge significantly affects the geometric accuracy and surface roughness in finish machining. A fuzzy estimator based on a fuzzy inference algorithm with a max-min composition rule is introduced to evaluate the minor flank wear length. The features sensitive to minor flank wear are extracted from the dispersion analysis of a time series AR model of the feed directional acceleration signal. These features, dispersions, are used for constructing linguistic rules, and then the fuzzy inferences are carried out with test data sets collected under various cutting conditions. The proposed system turns out to be effective for estimating minor flank wear length.

• Journal of Multimedia Information System
• /
• v.8 no.3
• /
• pp.147-158
• /
• 2021

Versatile Video Coding (VVC) is the latest video coding standard developed by Joint Video Exploration Team (JVET). In VVC, the quadtree plus multi-type tree (QT+MTT) structure of coding unit (CU) partition is adopted, and its computational complexity is considerably high due to the brute-force search for recursive rate-distortion (RD) optimization. In this paper, we aim to reduce the time complexity of inter-picture prediction mode since the inter prediction accounts for a large portion of the total encoding time. The problem can be defined as classifying the split mode of each CU. To classify the split mode effectively, a novel convolutional neural network (CNN) called multi-level tree (MLT-CNN) architecture is introduced. For boosting classification performance, we utilize additional information including inter-picture information while training the CNN. The overall algorithm including the MLT-CNN inference process is implemented on VVC Test Model (VTM) 11.0. The CUs of size 128×128 can be the inputs of the CNN. The sequences are encoded at the random access (RA) configuration with five QP values {22, 27, 32, 37, 42}. The experimental results show that the proposed algorithm can reduce the computational complexity by 11.53% on average, and 26.14% for the maximum with an average 1.01% of the increase in Bjøntegaard delta bit rate (BDBR). Especially, the proposed method shows higher performance on the sequences of the A and B classes, reducing 9.81%~26.14% of encoding time with 0.95%~3.28% of the BDBR increase.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.4
• /
• pp.402-410
• /
• 2004

In this paper, an optimal design method of clonal selection based Fuzzy-Neural Networks (FNN) model for complex and nonlinear systems is presented. The FNNs use the simplified inference as fuzzy inference method and Error Back Propagation Algorithm as learning rule. Also Advanced Clonal Selection (ACS) is proposed to find the parameters such as parameters of membership functions, learning rates and momentum coefficients. The proposed method is based on an Immune Algorithm (IA) using biological Immune System and The performance is improved by control of differentiation rate. Through that procedure, the antibodies are producted variously and the parameter of FNN are optimized by selecting method of antibody with the best affinity against antigens such as object function and limitation condition. To evaluate the performance of the proposed method, we use the time series data for gas furnace and traffic route choice process.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.5
• /
• pp.157-165
• /
• 2009

In this paper, we propose a distributed model that recognize ADLs of human can be occurred in daily living places. We collect and analyze user's environmental, location or activity information by simple sensor attached home devices or utensils. Based on these information, we provide a lifecare services by inferring the user's life pattern and health condition. But in order to provide a lifecare services well-refined activity recognition data are required and without enough inferred information it is very hard to build an ADL activity recognition model for high-level situation awareness. The sequence that generated by sensors are very helpful to infer the activities so we utilize the sequence to analyze an activity pattern and propose a distributed linear time inference algorithm. This algorithm is appropriate to recognize activities in small area like home, office or hospital. For performance evaluation, we test with an open data from MIT Media Lab and the recognition result shows over 75% accuracy.

• Journal of Electrical Engineering and Technology
• /
• v.5 no.2
• /
• pp.209-219
• /
• 2010

In this paper, a novel dynamic available transfer capability (DATC) has been computed for real time applications using three different intelligent techniques viz. i) back propagation algorithm (BPA), ii) radial basis function (RBF), and iii) adaptive neuro fuzzy inference system (ANFIS) for the first time. The conventional method of DATC is tedious and time consuming. DATC is concerned with calculating the maximum increase in point to point transfer such that the transient response remains stable and viable. The ATC information is to be continuously updated in real time and made available to market participants through an internet based Open Access Same time Information System (OASIS). The independent system operator (ISO) evaluates the transaction in real time on the basis of DATC information. The dynamic contingency screening method [1] has been utilized and critical contingencies are selected for the computation of DATC using the energy function based potential energy boundary surface (PEBS) method. The PEBS based DATC has been utilized to generate patterns for the intelligent techniques. The three different intelligent methods are tested on New England 68-bus 16 machine and 39-bus 10 machine systems and results are compared with the conventional PEBS method.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.4E
• /
• pp.146-155
• /
• 2002

We present a new technique for achieving source separation when given only a single charmel recording. The main idea is based on exploiting the inherent time structure of sound sources by learning a priori sets of time-domain basis functions that encode the sources in a statistically efficient manner. We derive a learning algorithm using a maximum likelihood approach given the observed single charmel data and sets of basis functions. For each time point we infer the source parameters and their contribution factors. This inference is possible due to the prior knowledge of the basis functions and the associated coefficient densities. A flexible model for density estimation allows accurate modeling of the observation, and our experimental results exhibit a high level of separation performance for simulated mixtures as well as real environment recordings employing mixtures of two different sources. We show separation results of two music signals as well as the separation of two voice signals.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.4
• /
• pp.146-146
• /
• 2002

We present a new technique for achieving source separation when given only a single channel recording. The main idea is based on exploiting the inherent time structure of sound sources by learning a priori sets of time-domain basis functions that encode the sources in a statistically efficient manner. We derive a learning algorithm using a maximum likelihood approach given the observed single channel data and sets of basis functions. For each time point we infer the source parameters and their contribution factors. This inference is possible due to the prior knowledge of the basis functions and the associated coefficient densities. A flexible model for density estimation allows accurate modeling of the observation, and our experimental results exhibit a high level of separation performance for simulated mixtures as well as real environment recordings employing mixtures of two different sources. We show separation results of two music signals as well as the separation of two voice signals.

• International Journal of Internet, Broadcasting and Communication
• /
• v.11 no.4
• /
• pp.76-85
• /
• 2019

We introduce a vision-based object detection method for real-time video surveillance system in low-end edge computing environments. Recently, the accuracy of object detection has been improved due to the performance of approaches based on deep learning algorithm such as Region Convolutional Neural Network(R-CNN) which has two stage for inferencing. On the other hand, one stage detection algorithms such as single-shot detection (SSD) and you only look once (YOLO) have been developed at the expense of some accuracy and can be used for real-time systems. However, high-performance hardware such as General-Purpose computing on Graphics Processing Unit(GPGPU) is required to still achieve excellent object detection performance and speed. To address hardware requirement that is burdensome to low-end edge computing environments, We propose sub-frame analysis method for the object detection. In specific, We divide a whole image frame into smaller ones then inference them on Convolutional Neural Network (CNN) based image detection network, which is much faster than conventional network designed forfull frame image. We reduced its computationalrequirementsignificantly without losing throughput and object detection accuracy with the proposed method.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.22 no.3
• /
• pp.644-653
• /
• 1998

In resistance spot weld, the assurance of weld quality has been a long-standing problem. Since the weld nuggets if resustance spot welding form between the workpieces, visual detection of defects in usually impossible. Welding quality of resistance spot welding can be verified by non destructive and destructive inspections such as X-Ray inspection and testing of weld strength. But these tests, in addition to being time-consuming and costly, can entail risks due to sampling basis. The purpose of this study is the development of the monitoring system based on fuzzy inference, aimed at diagonosis of quality in resistance spot welding. The fuzzy inference system consists of fuzzy input variables, fuzzy membership functions and fuzzy rules. For inferring the welding quality(strength), the experimental data of the spot welding were acquired in various welding conditions with the monitoring system designed. Some fuzzy input variables-maximum, slop and difference values of electrode movement signals-were extracted from the experimental data. It was confirmed that the fuzzy inference values of strength have a .${\pm}$5% error in comparison with actual values for the selected welding conditions(9-10.5KA, 10-14 cycle, 250-300 $kg_f$). This monitoring system can be useful in improving the quality assurance and reliability of the resistance spot welding process.