• Journal of Korea Multimedia Society
• /
• v.24 no.10
• /
• pp.1319-1325
• /
• 2021

It is important to remove fog for accurate object recognition and detection during preprocessing because images taken in foggy adverse weather suffer from poor quality of images due to scattering and absorption of light, resulting in poor performance of various vision-based applications. This paper proposes an end-to-end deep learning-based single image de-fogging method using U-Net architecture. The loss function used in the algorithm is a loss function based on Mahalanobis distance with fog features, which solves the problem of domain shifts, and demonstrates superior performance by comparing qualitative and quantitative numerical evaluations with conventional methods. We also design it to generate fog through the VGG19 loss function and use it as the next training dataset.

• Honam Mathematical Journal
• /
• v.38 no.3
• /
• pp.613-623
• /
• 2016

As a generalization of the Fourier transform, the fractional Fourier transform plays an important role both in theory and in applications of signal processing. We present a new approach to reach a uniform sampling theorem in the shift-invariant spaces associated with the fractional Fourier transform domain.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.29 no.9
• /
• pp.14-22
• /
• 2015

TIn this paper, a centering method for an unfocused input character using the spherical domain system and the centering character to use the shift invariant feature for the recognition system is proposed. A system for recognition is implemented using the centroid method with coordinate average values, and the results of an above 78.14% average differential ratio for the character features were obtained. It is possible to extract the shift invariant feature using spherical transformation similar to the human eyeball. The proposed method, which is feature extraction using spherical coordinate transform and transformed extracted data, makes it possible to move the character to the center position of the input plane. Both digital and optical technologies are mixed using a spherical coordinate similar to the 3 dimensional human eyeball for the 2 dimensional plane format. In this paper, a centering character feature using the spherical domain is proposed for character recognition, and possibilities for the recognized possible character shape as well as calculating the differential ratio of the centered character using a centroid method are suggested.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.1C
• /
• pp.56-63
• /
• 2003

In this paper, a wavelet-based stereo matching algorithm to obtain an accurate disparity map in wavelet transformed domain by using a shift ability property, a modified wavelet transform, the similarities for their sub-bands, and a hierarchical structure is proposed. New approaches for stereo matching by lots of feature information are to utilize translation-variant results of the sub-bands in the wavelet transformed domain because they cannot literally expect translation invariance in a system based on convolution and sub-sampling. After the similarity matching for each sub-band, we can easily find optimal matched-points because the sub-bands appearance of the shifted signals is definitely different from that of the original signal with no shift.

• Journal of the Optical Society of Korea
• /
• v.16 no.2
• /
• pp.170-173
• /
• 2012

We present a method that enables dual threshold voltages in a vertical alignment liquid crystal device, through which the gamma shift can be reduced with no subsequent decrease in the contrast ratio. By forming polymer layers, the threshold voltage shift is accomplished through the utilization of the voltage drop effect. We expect that the proposed method can be applied to the conventional 4-domain mode in order to achieve an 8-domain mode without the need for complex driving schemes.

• ETRI Journal
• /
• v.35 no.5
• /
• pp.838-848
• /
• 2013

This paper presents a novel approach to automatically generate Korean multiword sentiment expressions by using a seed sentiment lexicon and a large-scale domain-specific corpus. A multiword sentiment expression consists of a seed sentiment word and its contextual words occurring adjacent to the seed word. The multiword sentiment expressions that are the focus of our study have a different polarity from that of the seed sentiment word. The automatically extracted multiword sentiment expressions show that 1) the contextual words should be defined as a part of a multiword sentiment expression in addition to their corresponding seed sentiment word, 2) the identified multiword sentiment expressions contain various indicators for polarity shift that have rarely been recognized before, and 3) the newly recognized shifters contribute to assigning a more accurate polarity value. The empirical result shows that the proposed approach achieves improved performance of the sentiment analysis system that uses an automatically generated lexicon.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1993.06a
• /
• pp.210-214
• /
• 1993

There are two general techniques to observe particle movements in fluid. One is the acoustic time domain correlation technique and another is the frequency domain Doppler-shift techniques. Both techniques were reviewed and mentioned their some merits and demerits in ocean wave and current observation. Some possible application of acoustic time domain correlation technique in ocean wind wave measurement was discussed.

• Journal of Information Display
• /
• v.8 no.3
• /
• pp.17-21
• /
• 2007

We introduce a new electrode structure for the patterned vertical alignment (PVA) mode, which has a lower color-shift at large viewing angle than the one-domain VA mode. By manipulating the period of electrode structure, we can make more multi-domains and use the existing fabrication processes without having to use additional materials.

• Journal of Preventive Medicine and Public Health
• /
• v.37 no.2
• /
• pp.182-189
• /
• 2004

Objectives : The objective of this study is to compare the circadian patterns of heart rate variability assessed by 24-hour ambulatory electrocardiographic (ECG) recordings during day shift and night shift among the workers in the 5 days-concecutive-12-hour shift in an automobile factory in Korea. Methods : The study population consisted 300 workers, who were randomly selected among the 8700 total workers in one car factory. To analyse circadian variation, the 24-hour ECG recordings (Marquette) were measured during day shift (08:00-20:00 h) and night shift (20:00-08:00 h). Analysis was performed for all time and frequency domain measures of HRV. 233 workers completed taking 24-hour ECG recordings. Results : This study shows that the 24 hourcircadian variation mainly follows work/sleep cycle rather than day/night cycle among shift workers. This study also shows that among the night shift, the circadian variation between work and sleep cycle decreased compared to the work/sleep cycle among day shift workers. All time and frequency domain parameters (except LF/HF ratio) show significantly different between work and sleep in the day shift and night shift. Conclusion : These changes in heart rate variability circadian rhythms reflect significant reductions in cardiac parasympathetic activity with the most marked reduction in normal vagal activity among the shift workers. Especially, it suggests the circadian rhytm has blunted among the night workers. The quantification of the circadian variation in HRV can be a surrogates of workers' potential health risk, as well as suggests possible mechanisms through which the shift works compromise workers' health.

• Journal of information and communication convergence engineering
• /
• v.21 no.2
• /
• pp.139-144
• /
• 2023

Engineers prefer deep neural networks (DNNs) for solving computer vision problems. However, DNNs pose two major problems. First, neural networks require large amounts of well-labeled data for training. Second, the covariate shift problem is common in computer vision problems. Domain adaptation has been proposed to mitigate this problem. Recent work on adversarial-learning-based unsupervised domain adaptation (UDA) has explained transferability and enabled the model to learn robust features. Despite this advantage, current methods do not guarantee the distinguishability of the latent space unless they consider class-aware information of the target domain. Furthermore, source and target examples alone cannot efficiently extract domain-invariant features from the encoded spaces. To alleviate the problems of existing UDA methods, we propose the mixup regularization in adversarial discriminative domain adaptation (ADDA) method. We validated the effectiveness and generality of the proposed method by performing experiments under three adaptation scenarios: MNIST to USPS, SVHN to MNIST, and MNIST to MNIST-M.