• The Journal of Engineering Geology
• /
• v.32 no.3
• /
• pp.377-388
• /
• 2022

Recently, several studies have utilized machine learning to efficiently and accurately analyze seismic data that are exponentially increasing. In this study, we expand earthquake information such as occurrence time, hypocentral location, and magnitude to produce a dataset for applying to machine learning, reducing the dimension of the expended data into dominant features through principal component analysis. The dimensional extended data comprises statistics of the earthquake information from the Global Centroid Moment Tensor catalog containing 36,699 seismic events. We perform data preprocessing using standard and max-min scaling and extract dominant features with principal components analysis from the scaled dataset. The scaling methods significantly reduced the deviation of feature values caused by different units. Among them, the standard scaling method transforms the median of each feature with a smaller deviation than other scaling methods. The six principal components extracted from the non-scaled dataset explain 99% of the original data. The sixteen principal components from the datasets, which are applied with standardization or max-min scaling, reconstruct 98% of the original datasets. These results indicate that more principal components are needed to preserve original data information with even distributed feature values. We propose a data processing method for efficient and accurate machine learning model to analyze the relationship between seismic data and seismic behavior.

• The Journal of Bigdata
• /
• v.8 no.1
• /
• pp.157-165
• /
• 2023

Seafood is a major source of protein in many countries and its consumption is increasing. In Korea, consumption of seafood is increasing, but self-sufficiency rate is decreasing, and the importance of safety management is increasing as the amount of imported seafood increases. There are hundreds of species of aquatic products imported into Korea from over 110 countries, and there is a limit to relying only on the experience of inspectors for safety management of imported aquatic products. Based on the data, a model that can predict the customs inspection results of imported aquatic products is developed, and a machine learning classification model that determines the non-conformity of aquatic products when an import declaration is submitted is created. As a result of customs inspection of imported marine products, the nonconformity rate is less than 1%, which is very low imbalanced data. Therefore, a sampling method that can complement these characteristics was comparatively studied, and a preprocessing method that can interpret the classification result was applied. Among various machine learning-based classification models, Random Forest and XGBoost showed good performance. The model that predicts both compliance and non-conformance well as a result of the clearance inspection is the basic random forest model to which ADASYN and one-hot encoding are applied, and has an accuracy of 99.88%, precision of 99.87%, recall of 99.89%, and AUC of 99.88%. XGBoost is the most stable model with all indicators exceeding 90% regardless of oversampling and encoding type.

• Smart Media Journal
• /
• v.12 no.1
• /
• pp.9-16
• /
• 2023

Recently, a convolutional neural network (CNN) based system is being developed to overcome the limitations of human resources in the apple quality classification of farmhouse. However, since convolutional neural networks receive only images of the same size, preprocessing such as sampling may be required, and in the case of oversampling, information loss of the original image such as image quality degradation and blurring occurs. In this paper, in order to minimize the above problem, to generate a image patch based graph of an original image and propose a random walk-based positional encoding method to apply the graph transformer model. The above method continuously learns the position embedding information of patches which don't have a positional information based on the random walk algorithm, and finds the optimal graph structure by aggregating useful node information through the self-attention technique of graph transformer model. Therefore, it is robust and shows good performance even in a new graph structure of random node order and an arbitrary graph structure according to the location of an object in an image. As a result, when experimented with 5 apple quality datasets, the learning accuracy was higher than other GNN models by a minimum of 1.3% to a maximum of 4.7%, and the number of parameters was 3.59M, which was about 15% less than the 23.52M of the ResNet18 model. Therefore, it shows fast reasoning speed according to the reduction of the amount of computation and proves the effect.

• Geophysics and Geophysical Exploration
• /
• v.26 no.4
• /
• pp.171-184
• /
• 2023

Machine learning (ML)-based cavity detection using a large amount of survey data obtained from vehicle-mounted ground penetrating radar (GPR) has been actively studied to identify underground cavities. However, only simple image processing techniques have been used for preprocessing the ML input, and many conventional seismic and GPR data processing techniques, which have been used for decades, have not been fully exploited. In this study, based on the idea that a cavity can be identified using diffraction, we applied ML-based diffraction separation to GPR data to increase the accuracy of cavity detection using the YOLO v5 model. The original ML-based seismic diffraction separation technique was modified, and the separated diffraction image was used as the input to train the cavity detection model. The performance of the proposed method was verified using public GPR data released by the Seoul Metropolitan Government. Underground cavities and objects were more accurately detected using separated diffraction images. In the future, the proposed method can be useful in various fields in which GPR surveys are used.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.6
• /
• pp.250-258
• /
• 2007

Current speech recognition technology s achieved high performance with the development of hardware devices, however it is insufficient for some applications where high reliability is required, such as voice control of powered wheelchairs for disabled persons. For the system which aims to operate powered wheelchairs safely by voice in real environment, we need to consider that non-voice commands such as user s coughing, breathing, and spark-like mechanical noise should be rejected and the wheelchair system need to recognize the speech commands affected by disability, which contains specific pronunciation speed and frequency. In this paper, we propose non-voice rejection method to perform voice/non-voice classification using both YIN based fundamental frequency(F0) extraction and reliability in preprocessing. We adopted a multi-template dictionary and acoustic modeling based speaker adaptation to cope with the pronunciation variation of inarticulately uttered speech. From the recognition tests conducted with the data collected in real environment, proposed YIN based fundamental extraction showed recall-precision rate of 95.1% better than that of 62% by cepstrum based method. Recognition test by a new system applied with multi-template dictionary and MAP adaptation also showed much higher accuracy of 99.5% than that of 78.6% by baseline system.

• KIPS Transactions on Computer and Communication Systems
• /
• v.13 no.1
• /
• pp.21-30
• /
• 2024

As the self-driving car market continues to grow, the need for charging infrastructure is growing. However, in the case of a wireless charging system, stability issues are being raised because it requires a large amount of power compared with conventional wired charging. SAE J2954 is a standard for building autonomous vehicle wireless charging infrastructure, and the standard defines a communication method between a vehicle and a power transmission system. SAE J2954 recommends using physical media such as Wi-Fi, Bluetooth, and UWB as a wireless charging communication method for autonomous vehicles to enable communication between the vehicle and the charging pad. In particular, UWB is a suitable solution for indoor and outdoor charging environments because it exhibits robust communication capabilities in indoor environments and is not sensitive to interference. In this standard, the process for building a wireless power transmission system is divided into several stages from the start to the completion of charging. In this study, UWB technology is used as a means of fine alignment, a process in the wireless power transmission system. To determine the applicability to an actual autonomous vehicle wireless power transmission system, experiments were conducted based on distance, and the distance information was collected from UWB. To improve the accuracy of the distance data obtained from UWB, we propose a Single Model and Multi Model that apply machine learning and deep learning techniques to the collected data through a three-step preprocessing process.

• Progress in Medical Physics
• /
• v.24 no.2
• /
• pp.92-98
• /
• 2013

The experimental verification of treatment planning on the treatment spot is the ultimate method to assure quality of radiotherapy, so in-vivo skin dose measurement is the essential procedure to confirm treatment dose. In this study, glass rod dosimeter (GRD), which is a kind of photo-luminescent based dosimeters, was studied to produce a guideline to use GRDs in vivo dosimetry for quality assurance of radiotherapy. The pre-processing procedure is essential to use GRDs. This is a heating operation for stabilization. Two kinds of pre-processing methods are recommended by manufacturer: a heating method (70 degree, 30 minutes) and a waiting method (room temperature, 24 hours). We equally irradiated 1.0 Gy to 20 GRD elements, and then different preprocessing were performed to 10 GRDs each. In heating method, reading deviation of GRDs at same time were relatively high, but the deviation was very low as time went on. In waiting method, the deviation among GRDs was low, but the deviation was relatively high as time went on. The meaningful difference was found between mean reading values of two pre-processing methods. Both methods present mean dose deviation under 5%, but the relatively high effect by reading time was observed in waiting method. Finally, GRD is best to perform in-vivo dosimetry in the viewpoint of accuracy and efficiency, and the understanding of how pre-processing affect the accuracy is asked to perform most accurate in-vivo dosimetry. The further study is asked to acquire more stable accuracy in spite of different irradiation conditions for GRD usage.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.189-198
• /
• 2024

Lipreading is one of the important parts of speech recognition, and several studies have been conducted to improve the performance of lipreading in lipreading systems for speech recognition. Recent studies have used method to modify the model architecture of lipreading system to improve recognition performance. Unlike previous research that improve recognition performance by modifying model architecture, we aim to improve recognition performance without any change in model architecture. In order to improve the recognition performance without modifying the model architecture, we refer to the cues used in human lipreading and set other regions such as chin and cheeks as regions of interest along with the lip region, which is the existing region of interest of lipreading systems, and compare the recognition rate of each region of interest to propose the highest performing region of interest In addition, assuming that the difference in normalization results caused by the difference in interpolation method during the process of normalizing the size of the region of interest affects the recognition performance, we interpolate the same region of interest using nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation, and compare the recognition rate of each interpolation method to propose the best performing interpolation method. Each region of interest was detected by training an object detection neural network, and dynamic time warping templates were generated by normalizing each region of interest, extracting and combining features, and mapping the dimensionality reduction of the combined features into a low-dimensional space. The recognition rate was evaluated by comparing the distance between the generated dynamic time warping templates and the data mapped to the low-dimensional space. In the comparison of regions of interest, the result of the region of interest containing only the lip region showed an average recognition rate of 97.36%, which is 3.44% higher than the average recognition rate of 93.92% in the previous study, and in the comparison of interpolation methods, the bilinear interpolation method performed 97.36%, which is 14.65% higher than the nearest neighbor interpolation method and 5.55% higher than the bicubic interpolation method. The code used in this study can be found a https://github.com/haraisi2/Lipreading-Systems.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.97-117
• /
• 2020

Product evaluation criteria is an indicator describing attributes or values of products, which enable users or manufacturers measure and understand the products. When companies analyze their products or compare them with competitors, appropriate criteria must be selected for objective evaluation. The criteria should show the features of products that consumers considered when they purchased, used and evaluated the products. However, current evaluation criteria do not reflect different consumers' opinion from product to product. Previous studies tried to used online reviews from e-commerce sites that reflect consumer opinions to extract the features and topics of products and use them as evaluation criteria. However, there is still a limit that they produce irrelevant criteria to products due to extracted or improper words are not refined. To overcome this limitation, this research suggests LDA-k-NN model which extracts possible criteria words from online reviews by using LDA and refines them with k-nearest neighbor. Proposed approach starts with preparation phase, which is constructed with 6 steps. At first, it collects review data from e-commerce websites. Most e-commerce websites classify their selling items by high-level, middle-level, and low-level categories. Review data for preparation phase are gathered from each middle-level category and collapsed later, which is to present single high-level category. Next, nouns, adjectives, adverbs, and verbs are extracted from reviews by getting part of speech information using morpheme analysis module. After preprocessing, words per each topic from review are shown with LDA and only nouns in topic words are chosen as potential words for criteria. Then, words are tagged based on possibility of criteria for each middle-level category. Next, every tagged word is vectorized by pre-trained word embedding model. Finally, k-nearest neighbor case-based approach is used to classify each word with tags. After setting up preparation phase, criteria extraction phase is conducted with low-level categories. This phase starts with crawling reviews in the corresponding low-level category. Same preprocessing as preparation phase is conducted using morpheme analysis module and LDA. Possible criteria words are extracted by getting nouns from the data and vectorized by pre-trained word embedding model. Finally, evaluation criteria are extracted by refining possible criteria words using k-nearest neighbor approach and reference proportion of each word in the words set. To evaluate the performance of the proposed model, an experiment was conducted with review on '11st', one of the biggest e-commerce companies in Korea. Review data were from 'Electronics/Digital' section, one of high-level categories in 11st. For performance evaluation of suggested model, three other models were used for comparing with the suggested model; actual criteria of 11st, a model that extracts nouns by morpheme analysis module and refines them according to word frequency, and a model that extracts nouns from LDA topics and refines them by word frequency. The performance evaluation was set to predict evaluation criteria of 10 low-level categories with the suggested model and 3 models above. Criteria words extracted from each model were combined into a single words set and it was used for survey questionnaires. In the survey, respondents chose every item they consider as appropriate criteria for each category. Each model got its score when chosen words were extracted from that model. The suggested model had higher scores than other models in 8 out of 10 low-level categories. By conducting paired t-tests on scores of each model, we confirmed that the suggested model shows better performance in 26 tests out of 30. In addition, the suggested model was the best model in terms of accuracy. This research proposes evaluation criteria extracting method that combines topic extraction using LDA and refinement with k-nearest neighbor approach. This method overcomes the limits of previous dictionary-based models and frequency-based refinement models. This study can contribute to improve review analysis for deriving business insights in e-commerce market.

• Journal of Broadcast Engineering
• /
• v.22 no.2
• /
• pp.193-206
• /
• 2017

In this paper, we propose a rate control method based on the $R-{\lambda}$ model that supports a parallel encoding structure in GOP levels or IDR period levels for 4K UHD input video in real-time. For this, a slice-level bit allocation method is proposed for parallel encoding instead of sequential encoding. When a rate control algorithm is applied in the GOP level or IDR period level parallelism, the information of how many bits are consumed cannot be shared among the frames belonging to a same frame level except the lowest frame level of the hierarchical B structure. Therefore, it is impossible to manage the bit budget with the existing bit allocation method. In order to solve this problem, we improve the bit allocation procedure of the conventional ones that allocate target bits sequentially according to the encoding order. That is, the proposed bit allocation strategy is to assign the target bits in GOPs first, then to distribute the assigned target bits from the lowest depth level to the highest depth level of the HEVC hierarchical B structure within each GOP. In addition, we proposed a processing method that is used to improve subjective image qualities by allocating the bits according to the coding complexities of the frames. Experimental results show that the proposed bit allocation method works well for frame-level parallel HEVC software encoders and it is confirmed that the performance of our rate controller can be improved with a more elaborate bit allocation strategy by using the preprocessing results.