• Journal of Ocean Engineering and Technology
• /
• v.31 no.1
• /
• pp.69-79
• /
• 2017

This paper presents the operations of a multi-beam echo sounder (MBES) installed on the deep-sea remotely operated vehicle (ROV) Hemire. Hemire explored hydrothermal vents in the Forecast volcano located near the Mariana Trench in March of in 2006. During these explorations, we acquired profiling points on the routes of the vehicle using the MBES. Information on the position, depth, and attitude of the ROV are essential to obtain higher accuracy for the profiling quality. However, the MBES installed on Hemire does not have its own position and depth sensors. Although it has attitude sensors for roll, pitch, and heading, the specifications of these sensors were not clear. Therefore, we had to merge the high-performance sensor data for the motion and position obtained from Hemire into the profiling data of the MBES. Then, we could properly convert the profiling points with respect to the Earth-fixed coordinates. This paper describes the integration of the MBES with Hemire, as well as the coordinate conversion between them. Bathymetric maps near the summit of the Forecast volcano were successfully collected through these processes. A comparison between the bathymetric maps from the MBES and those from the Onnuri Research Vessel, the mother ship of the ROV Hemire for these explorations, is also presented.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.5
• /
• pp.193-204
• /
• 2019

The explosion of data due to the improvement of sensor technology and computing performance has become the basis for analyzing the situation in the industrial fields, and various attempts to detect events based on such data are increasing recently. In particular, sound signals collected from sensors are used as important information to classify events in various application fields as an advantage of efficiently collecting field information at a relatively low cost. However, the performance of sound-event classification in the field cannot be guaranteed if noise can not be removed. That is, in order to implement a system that can be practically applied, robust performance should be guaranteed even in various noise conditions. In this study, we propose a system that can classify the sound event after generating the enhanced sound signal based on the deep learning algorithm. Especially, to remove noise from the sound signal itself, the enhanced sound data against the noise is generated using SEGAN applied to the GAN with a VAE technique. Then, an end-to-end based sound-event classification system is designed to classify the sound events using the enhanced sound signal as input data of CNN structure without a data conversion process. The performance of the proposed method was verified experimentally using sound data obtained from the industrial field, and the f1 score of 99.29% (railway industry) and 97.80% (livestock industry) was confirmed.

• Journal of Internet Computing and Services
• /
• v.23 no.4
• /
• pp.35-43
• /
• 2022

Makeup is the most common way to improve a person's appearance. However, since makeup styles are very diverse, there are many time and cost problems for an individual to apply makeup directly to himself/herself.. Accordingly, the need for makeup automation is increasing. Makeup transfer is being studied for makeup automation. Makeup transfer is a field of applying makeup style to a face image without makeup. Makeup transfer can be divided into a traditional image processing-based method and a deep learning-based method. In particular, in deep learning-based methods, many studies based on Generative Adversarial Networks have been performed. However, both methods have disadvantages in that the resulting image is unnatural, the result of makeup conversion is not clear, and it is smeared or heavily influenced by the makeup style face image. In order to express the clear boundary of makeup and to alleviate the influence of makeup style facial images, this study divides the makeup area and calculates the loss function using HoG (Histogram of Gradient). HoG is a method of extracting image features through the size and directionality of edges present in the image. Through this, we propose a makeup transfer network that performs robust learning on edges.By comparing the image generated through the proposed model with the image generated through BeautyGAN used as the base model, it was confirmed that the performance of the model proposed in this study was superior, and the method of using facial information that can be additionally presented as a future study.

• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.23-29
• /
• 2021

Recent deep learning-based natural language processing studies are conducting research to improve performance by training large amounts of data from various sources together. However, there is a possibility that the methodology of learning by combining data from various sources into one may prevent performance improvement. In the case of machine translation, data deviation occurs due to differences in translation(liberal, literal), style(colloquial, written, formal, etc.), domains, etc. Combining these corpora into one for learning can adversely affect performance. In this paper, we propose a new Corpus Weight Balance(CWB) method that considers the balance between parallel corpora in machine translation. As a result of the experiment, the model trained with balanced corpus showed better performance than the existing model. In addition, we propose an additional corpus construction process that enables coexistence with the human translation market, which can build high-quality parallel corpus even with a monolingual corpus.

• Geophysics and Geophysical Exploration
• /
• v.26 no.3
• /
• pp.138-149
• /
• 2023

In geoscience and engineering the geological characteristics of sediment strata is crucial and possible if reliable borehole logging and seismic data are available. To investigate the characteristics of the shallow strata in the Korea Strait, laboratory sonic logs were obtained from deep borehole data and seismic section. In this study, we integrated and analyzed the sonic log data obtained from the drilling core (down to a depth of 200 m below the seabed) and multichannel seismic section. The correlation value was increased from 15% to 45% through time-depth conversion. An initial model of P-wave impedance was set, and the results were compared by performing model-based, band-limited, and sparse-spike inversions. The derived P-impedance distributions exhibited differences between sediment-dominant and unconsolidated layers. The P-impedance inversion process can be used as a framework for an integrated analysis of additional core logs and seismic data in the future. Furthermore, the derived P-impedance can be used to detect shallow gas-saturated regions or faults in the shallow sediment. As domestic deep drilling is being performed continuously for identifying the characteristics of carbon dioxide storage candidates and evaluating resources, the applicability of the integrated inversion will increase in the future.

• Journal of Intelligence and Information Systems
• /
• v.24 no.2
• /
• pp.221-241
• /
• 2018

Deep learning is getting attention recently. The deep learning technique which had been applied in competitions of the International Conference on Image Recognition Technology(ILSVR) and AlphaGo is Convolution Neural Network(CNN). CNN is characterized in that the input image is divided into small sections to recognize the partial features and combine them to recognize as a whole. Deep learning technologies are expected to bring a lot of changes in our lives, but until now, its applications have been limited to image recognition and natural language processing. The use of deep learning techniques for business problems is still an early research stage. If their performance is proved, they can be applied to traditional business problems such as future marketing response prediction, fraud transaction detection, bankruptcy prediction, and so on. So, it is a very meaningful experiment to diagnose the possibility of solving business problems using deep learning technologies based on the case of online shopping companies which have big data, are relatively easy to identify customer behavior and has high utilization values. Especially, in online shopping companies, the competition environment is rapidly changing and becoming more intense. Therefore, analysis of customer behavior for maximizing profit is becoming more and more important for online shopping companies. In this study, we propose 'CNN model of Heterogeneous Information Integration' using CNN as a way to improve the predictive power of customer behavior in online shopping enterprises. In order to propose a model that optimizes the performance, which is a model that learns from the convolution neural network of the multi-layer perceptron structure by combining structured and unstructured information, this model uses 'heterogeneous information integration', 'unstructured information vector conversion', 'multi-layer perceptron design', and evaluate the performance of each architecture, and confirm the proposed model based on the results. In addition, the target variables for predicting customer behavior are defined as six binary classification problems: re-purchaser, churn, frequent shopper, frequent refund shopper, high amount shopper, high discount shopper. In order to verify the usefulness of the proposed model, we conducted experiments using actual data of domestic specific online shopping company. This experiment uses actual transactions, customers, and VOC data of specific online shopping company in Korea. Data extraction criteria are defined for 47,947 customers who registered at least one VOC in January 2011 (1 month). The customer profiles of these customers, as well as a total of 19 months of trading data from September 2010 to March 2012, and VOCs posted for a month are used. The experiment of this study is divided into two stages. In the first step, we evaluate three architectures that affect the performance of the proposed model and select optimal parameters. We evaluate the performance with the proposed model. Experimental results show that the proposed model, which combines both structured and unstructured information, is superior compared to NBC(Naïve Bayes classification), SVM(Support vector machine), and ANN(Artificial neural network). Therefore, it is significant that the use of unstructured information contributes to predict customer behavior, and that CNN can be applied to solve business problems as well as image recognition and natural language processing problems. It can be confirmed through experiments that CNN is more effective in understanding and interpreting the meaning of context in text VOC data. And it is significant that the empirical research based on the actual data of the e-commerce company can extract very meaningful information from the VOC data written in the text format directly by the customer in the prediction of the customer behavior. Finally, through various experiments, it is possible to say that the proposed model provides useful information for the future research related to the parameter selection and its performance.

• Asian-Australasian Journal of Animal Sciences
• /
• v.27 no.7
• /
• pp.974-980
• /
• 2014

In semi-arid areas of Southern Africa, dehydration can compromise the performance and welfare of local chickens, particularly during the growing period when confinement is curtailed and birds are left to scavenge for feed and water. The effect of water restriction on the growth performance was compared in Naked Neck (NNK) and Ovambo (OVB) chickens that are predominant in Southern Africa. A total of 54 eight-wk-old pullets each of NNK and OVB chickens with an initial average weight of $641{\pm}10g/bird$ were randomly assigned to three water intake treatments, each having six birds for 8 wk. The water restriction treatments were ad libitum, 70% of ad libitum and 40% of ad libitum intake. Nine experimental pens with a floor space of $3.3m^2$ per strain were used. The pens were housed in an open-sided house with cement floor deep littered with a 20 cm layer of untreated wood shavings. Feed was provided ad libitum. Average daily water intake (ADWI), BW at 16 weeks of age (FBW), ADG, ADFI, feed conversion ratio (FCR) and water to feed ratios (WFR) were determined. Ovambo chickens had superior (p<0.05) FBW, ADG and ADWI than NNK chickens. Body weight of birds at 16 weeks of age, ADG, ADFI, ADWI, and WFR declined progressively (p<0.05) with increasing severity of water restriction while FCR values increased (p<0.05) as the severity of water restriction increased. Naked Neck chickens had better FCR at the 40% of ad libitum water intake level than Ovambo chickens. The dressing percentage per bird was higher in water restricted birds than those on ad libitum water consumption, irrespective of strain. Heart weight was significantly lower in birds on 40% of ad libitum water intake than those on ad libitum and 70% of ad libitum water intake, respectively. In conclusion, NNK chickens performed better than OVB chickens under conditions of water restriction and would be ideal to raise for meat and egg production in locations where water shortages are a major challenge.

• Asian-Australasian Journal of Animal Sciences
• /
• v.14 no.5
• /
• pp.677-679
• /
• 2001

Effect of Growell a herbomineral premix was evaluated on performance, organ weight, humoral immune response, tissue and serum trace element profile of boilers. Day old 50 Babcock broiler chicks were randomly divided in two groups (C and T) and reared on deep litter system for 6 weeks of age. Chicks from group C were given basal diet while chicks from T group were fed basal diet supplemented with Growell at the rate 0.35 g/Kg. The chicks were vaccinated with Lasota strain of NCDV at $4^{th}$ and $28^{th}$ day of age. The birds receiving Growell treatment had higher body weight with better feed conversion ratio as compare to that of control chicks. Growell treatment had significantly reduced per cent weight of spleen and kidney, whereas that of bursa was increased. There was no effect of treatment on relative weight of thymus. HI antibody titer against NCDV in Growell treated chicks were higher as compare to untreated chicks indicating better humoral immune status. Growell treatment had no effect on serum Fe and Zn concentration. Dietary supplementation of Growell had significantly increased iron content of liver, kidney and muscle; zinc content of kidney and muscle; copper content of kidney and muscle and Mn content of kidney. Growell treatment improved the body weight, FCR and humoral immune status of broilers. Similarly, deposition of trace minerals in various organs was also increased in comparison to control.

• Journal of Internet of Things and Convergence
• /
• v.4 no.2
• /
• pp.1-6
• /
• 2018

Techniques for generating new sample data from higher dimensional data such as images have been utilized variously for speech synthesis, image conversion and image restoration. This paper adopts Progressive Growing of Generative Adversarial Networks(PG-GANs) as an implementation model to generate high-resolution images and to enhance variation of the generated images, and applied it to fashion image data. PG-GANs allows the generator and discriminator to progressively learn at the same time, continuously adding new layers from low-resolution images to result high-resolution images. We also proposed a Mini-batch Discrimination method to increase the diversity of generated data, and proposed a Sliced Wasserstein Distance(SWD) evaluation method instead of the existing MS-SSIM to evaluate the GAN model.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.17 no.1
• /
• pp.129-140
• /
• 2017

In this paper, a $4{\times}32$-channel neural recording system capable of acquiring neural signals is introduced. Four 32-channel neural recording ICs, complex programmable logic devices (CPLDs), a micro controller unit (MCU) with USB interface, and a PC are used. Each neural recording IC, implemented in $0.18{\mu}m$ CMOS technology, includes 32 channels of analog front-ends (AFEs), a 32-to-1 analog multiplexer, and an analog-to-digital converter (ADC). The mid-band gain of the AFE is adjustable in four steps, and have a tunable bandwidth. The AFE has a mid-band gain of 54.5 dB to 65.7 dB and a bandwidth of 35.3 Hz to 5.8 kHz. The high-pass cutoff frequency of the AFE varies from 18.6 Hz to 154.7 Hz. The input-referred noise (IRN) of the AFE is $10.2{\mu}V_{rms}$. A high-resolution, low-power ADC with a high conversion speed achieves a signal-to-noise and distortion ratio (SNDR) of 50.63 dB and a spurious-free dynamic range (SFDR) of 63.88 dB, at a sampling-rate of 2.5 MS/s. The effectiveness of our neural recording system is validated in in-vivo recording of the primary somatosensory cortex of a rat.