• Journal of KIISE:Software and Applications
• /
• v.34 no.12
• /
• pp.1045-1055
• /
• 2007

Latent topic models are statistical models which automatically captures salient patterns or correlation among features underlying a data collection in a probabilistic way. They are gaining an increased popularity as an effective tool in the application of automatic semantic feature extraction from text corpus, multimedia data analysis including image data, and bioinformatics. Among the important issues for the effectiveness in the application of latent topic models to the massive data set is the efficient learning of the model. The paper proposes an accelerated learning technique for PLSA model, one of the popular latent topic models, by an incremental EM algorithm instead of conventional EM algorithm. The incremental EM algorithm can be characterized by the employment of a series of partial E-steps that are performed on the corresponding subsets of the entire data collection, unlike in the conventional EM algorithm where one batch E-step is done for the whole data set. By the replacement of a single batch E-M step with a series of partial E-steps and M-steps, the inference result for the previous data subset can be directly reflected to the next inference process, which can enhance the learning speed for the entire data set. The algorithm is advantageous also in that it is guaranteed to converge to a local maximum solution and can be easily implemented just with slight modification of the existing algorithm based on the conventional EM. We present the basic application of the incremental EM algorithm to the learning of PLSA and empirically evaluate the acceleration performance with several possible data partitioning methods for the practical application. The experimental results on a real-world news data set show that the proposed approach can accomplish a meaningful enhancement of the convergence rate in the learning of latent topic model. Additionally, we present an interesting result which supports a possible synergistic effect of the combination of incremental EM algorithm with parallel computing.

• The KIPS Transactions:PartB
• /
• v.11B no.3
• /
• pp.281-288
• /
• 2004

In order to extract and analyze complex features of the behavior of animals in response to external stimuli such as toxic chemicals, we implemented an adaptive computational method to characterize changes in the behavior of chironomids in response to treatment with the insecticide, diazinon. In this paper, we propose an energy minimization model to extract the features of response behavior of chironomids under toxic treatment, which is applied on the image of velocity vectors. It is based on the improved active contour model and the variations of the energy functional, which are produced by the evolving active contour. The movement tracks of individual chironomid larvae were continuously measured in 0.25 second intervals during the survey period of 4 days before and after the treatment. Velocity on each sample track at 0.25 second intervals was collected in 15-20 minute periods and was subsequently checked to effectively reveal behavioral states of the specimens tested. Active contour was formed around each collection of velocities to gradually evolve to find the optimal boundaries of velocity collections through processes of energy minimization. The active contour which is improved by T. Chan and L. Vese is used in this paper. The energy minimization model effectively revealed characteristic patterns of behavior for the treatment versus no treatment, and identified changes in behavioral states .is the time progressed.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.12
• /
• pp.546-554
• /
• 2019

Typically, closed-circuit television (CCTV) monitoring is mainly used for post-processes (i.e. to provide evidence after an incident has occurred), but by using a streaming video feed, machine-based learning, and advanced image recognition techniques, current technology can be extended to respond to crimes or reports of missing persons in real time. The multi-CCTV cooperation technique developed in this study is a program model that delivers similarity information about a suspect (or moving object) extracted via CCTV at one location and sent to a monitoring agent to track the selected suspect or object when he, she, or it moves out of range to another CCTV camera. To improve the operating efficiency of local government CCTV control centers, we describe here the partial automation of a CCTV control system that currently relies upon monitoring by human agents. We envisage an integrated crime prevention service, which incorporates the cooperative CCTV network suggested in this study and that can easily be experienced by citizens in ways such as determining a precise individual location in real time and providing a crime prevention service linked to smartphones and/or crime prevention/safety information.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.6
• /
• pp.433-438
• /
• 2016

This paper proposes a grade prediction method to measure meat quality in Hanwoo (Korean Native Cattle) using classification and feature extraction algorithms. The applied classification algorithm is an AdaBoost and the texture features of the given ultrasound images are extracted using SFTA. In this paper, as an initial phase, we selected ultrasound images of Hanwoo for verifying experimental results; however, we ultimately aimed to develop a diagnostic decision support system for human body scan using ultrasound images. The advantages of using ultrasound images of Hanwoo are: accurate grade prediction without butchery, optimizing shipping and feeding schedule and economic benefits. Researches on grade prediction using biometric data such as ultrasound images have been studied in countries like USA, Japan, and Korea. Studies have been based on accurate prediction method of different images obtained from different machines. However, the prediction accuracy is low. Therefore, we proposed a prediction method of meat quality. From the experimental results compared with that of the real grades, the experimental results demonstrated that the proposed method is superior to the other methods.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.41-43
• /
• 2022

In order to prevent and block infectious diseases caused by the recent COVID-19 pandemic, non-contact biometric information acquisition and analysis technology is attracting attention. The invasive and attached biometric information acquisition method accurately has the advantage of measuring biometric information, but has a risk of increasing contagious diseases due to the close contact. To solve these problems, the non-contact method of extracting biometric information such as human fingerprints, faces, iris, veins, voice, and signatures with automated devices is increasing in various industries as data processing speed increases and recognition accuracy increases. However, although the accuracy of the non-contact biometric data acquisition technology is improved, the non-contact method is greatly influenced by the surrounding environment of the object to be measured, which is resulting in distortion of measurement information and poor accuracy. In this paper, we propose a context-based bio-signal modeling technique for the interpretation of personalized information (image, signal, etc.) for bio-information analysis. Context-based biometric information modeling techniques present a model that considers contextual and user information in biometric information measurement in order to improve performance. The proposed model analyzes signal information based on the feature probability distribution through context-based signal analysis that can maximize the predicted value probability.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.19 no.4
• /
• pp.63-75
• /
• 2016

Aquaculture has historically delivered marine products because the country is surrounded by ocean on three sides. Surveys on production have been conducted recently to systematically manage aquaculture facilities. Based on survey results, pricing controls on marine products has been implemented to stabilize local fishery resources and to ensure minimum income for fishermen. Such surveys on aquaculture facilities depend on manual digitization of aerial photographs each year. These surveys that incorporate manual digitization using high-resolution aerial photographs can accurately evaluate aquaculture with the knowledge of experts, who are aware of each aquaculture facility's characteristics and deployment of those facilities. However, using aerial photographs has monetary and time limitations for monitoring aquaculture resources with different life cycles, and also requires a number of experts. Therefore, in this study, we investigated an automatic prototype system for detecting boundary information and monitoring aquaculture facilities based on satellite images. KOMPSAT-3 (13 Scene), a local high-resolution satellite provided the satellite imagery collected between October and April, a time period in which many aquaculture facilities were operating. The ANN classification method was used for automatic detecting such as cage, longline and buoy type. Furthermore, shape files were generated using a digitizing image processing method that incorporates polygon generation techniques. In this study, our newly developed prototype method detected aquaculture facilities at a rate of 93%. The suggested method overcomes the limits of existing monitoring method using aerial photographs, but also assists experts in detecting aquaculture facilities. Aquaculture facility detection systems must be developed in the future through application of image processing techniques and classification of aquaculture facilities. Such systems will assist in related decision-making through aquaculture facility monitoring.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.175-197
• /
• 2021

Recently, with the development of deep learning technology, research on unstructured data analysis is being actively conducted, and it is showing remarkable results in various fields such as classification, summary, and generation. Among various text analysis fields, text classification is the most widely used technology in academia and industry. Text classification includes binary class classification with one label among two classes, multi-class classification with one label among several classes, and multi-label classification with multiple labels among several classes. In particular, multi-label classification requires a different training method from binary class classification and multi-class classification because of the characteristic of having multiple labels. In addition, since the number of labels to be predicted increases as the number of labels and classes increases, there is a limitation in that performance improvement is difficult due to an increase in prediction difficulty. To overcome these limitations, (i) compressing the initially given high-dimensional label space into a low-dimensional latent label space, (ii) after performing training to predict the compressed label, (iii) restoring the predicted label to the high-dimensional original label space, research on label embedding is being actively conducted. Typical label embedding techniques include Principal Label Space Transformation (PLST), Multi-Label Classification via Boolean Matrix Decomposition (MLC-BMaD), and Bayesian Multi-Label Compressed Sensing (BML-CS). However, since these techniques consider only the linear relationship between labels or compress the labels by random transformation, it is difficult to understand the non-linear relationship between labels, so there is a limitation in that it is not possible to create a latent label space sufficiently containing the information of the original label. Recently, there have been increasing attempts to improve performance by applying deep learning technology to label embedding. Label embedding using an autoencoder, a deep learning model that is effective for data compression and restoration, is representative. However, the traditional autoencoder-based label embedding has a limitation in that a large amount of information loss occurs when compressing a high-dimensional label space having a myriad of classes into a low-dimensional latent label space. This can be found in the gradient loss problem that occurs in the backpropagation process of learning. To solve this problem, skip connection was devised, and by adding the input of the layer to the output to prevent gradient loss during backpropagation, efficient learning is possible even when the layer is deep. Skip connection is mainly used for image feature extraction in convolutional neural networks, but studies using skip connection in autoencoder or label embedding process are still lacking. Therefore, in this study, we propose an autoencoder-based label embedding methodology in which skip connections are added to each of the encoder and decoder to form a low-dimensional latent label space that reflects the information of the high-dimensional label space well. In addition, the proposed methodology was applied to actual paper keywords to derive the high-dimensional keyword label space and the low-dimensional latent label space. Using this, we conducted an experiment to predict the compressed keyword vector existing in the latent label space from the paper abstract and to evaluate the multi-label classification by restoring the predicted keyword vector back to the original label space. As a result, the accuracy, precision, recall, and F1 score used as performance indicators showed far superior performance in multi-label classification based on the proposed methodology compared to traditional multi-label classification methods. This can be seen that the low-dimensional latent label space derived through the proposed methodology well reflected the information of the high-dimensional label space, which ultimately led to the improvement of the performance of the multi-label classification itself. In addition, the utility of the proposed methodology was identified by comparing the performance of the proposed methodology according to the domain characteristics and the number of dimensions of the latent label space.