This paper proposes an image classification algorithm that transforms the number of convolution layers in the residual block of ResNet, CNN's representative method. The proposed method modified the structure of 34/50 layer of ResNet structure. First, we analyzed the performance of small and many convolution layers for the structure consisting of only shortcut and 3 × 3 convolution layers for 34 and 50 layers. And then the performance was analyzed in the case of small and many cases of convolutional layers for the bottleneck structure of 50 layers. By applying the results, the best classification method in the residual block was applied to construct a 34-layer simple structure and a 50-layer bottleneck image classification model. To evaluate the performance of the proposed image classification model, the results were analyzed by applying to the cifar10 dataset. The proposed 34-layer simple structure and 50-layer bottleneck showed improved performance over the ResNet-110 and Densnet-40 models.

Cervical cancer is the second most common female cancer in the world. In Korea, cervical cancer accounts for 13 percent of female cancers and 4,200 cases occur annually[1]. The purpose of this study is to use a deep learning model to identify the possibility of lesions in the cervix and to evaluate the efficient image preprocessing in order to diagnose diverse types of cervix in form. The study used 4,107 normal photographs of uterine cervix and 6,285 abnormal photographs of uterine cervix. Two types of image preprocessing were resized to square. The methods are cropping based on height and filling the space up and down with black images. In addition, all images were resampled to 256×256. The average accuracy of cropped cases is 94.15%. The average accuracy of the filled cases is 93.41%. According to the study, the model performance of cropped data was slightly better. But there were several images that were not accurately classified. Therefore, the additional experiment with pre-treatment process based on cropping is needed to cover images of the cervix in more detail.

Underwater communication is necessarily useful for various application domains such as saving of human lives from underwater disasters, marine resource exploration, underwater military fields, underwater environment or ecosystem monitoring, fish farm monitoring and management, etc. Even though the acoustic wave has been the main underwater communication media until now, several media such as optical waves, VLF/ELF waves, magnetic fields, and infrared rays also began to be treated as possible media for underwater communication. If these underwater communicate-possible media are used mixing together, the underwater communication can be much more reliable and efficient through complementing the disadvantages of each communication media with advantages of other communication media. In fact, mixing and using multi-media for underwater communication requires the data flow control in the connection process of different media due to their communication speed gaps and bandwidth differences, and, specially, in the flow control, the appropriate message fragmentation technique is required inevitably. For this reason, this paper presents a fragmentation framework and technique necessary to the flow control in the underwater multi-media communication. In addition, through its implementation and experiments, this paper shows the feasibility on the realization of the multi-media based underwater communication.

As a popular mathematical framework for modeling decision making, Markov decision process (MDP) has been widely used to solve problem in many engineering fields. MDP consists of a set of discrete states, a finite set of actions, and rewards received after reaching a new state by taking action from the previous state. The objective of MDP is to find an optimal policy, that is, to find the best action to be taken in each state to maximize the expected discounted reward of policy (EDR). In practice, MDP is typically unknown, so simulation-based policy improvement (SBPI), which improves a given base policy sequentially by selecting the best action in each state depending on rewards observed via simulation, can be a practical way to find the optimal policy. However, the efficiency of SBPI is still a concern since many simulation samples are required to precisely estimate EDR for each action in each state. In this paper, we propose a method to select the best action accurately in each state using a small number of simulation samples, thereby improving the efficiency of SBPI. The proposed method accumulates the simulation samples observed in the previous states, so it is possible to precisely estimate EDR even with a small number of samples in the current state. The results of comparative experiments on the existing method demonstrate that the proposed method can improve the efficiency of SBPI.

Walt Disney Feature Animation released high quality films such as 'Frozen', 'Big Hero 6', 'Wreck-it Ralph', 'Zootopia', 'Moana', 'Frozen 2' and not only got high score in box office but showed great CG and visuals. However, making feature animation requires a lot of time, money, and efforts so it is very important to support studios' artists to finish each show within limited budget and time. This paper shows artists' training program such as 'Short Circuit' and 'Bootcamp' that walt disney feature animation provides their artists to improve their creativity and do their jobs artistically and efficiently. Disney's training program not only provides artists various training classes but gives them chances to work on short animation which enhances artistic skills and enable them to work in different departments and experience different tasks. This paper also explains some training cases of CG studios in South Korea and Disney Animations' in-house tools.

Hangul is a language that is composed of initial, medial, and final syllables. It has 11,172 characters. For this reason, the current method of designing all the characters by hand is very expensive and time-consuming. In order to solve the problem, this paper proposes an automatic Hangul font generation system and evaluates the standards for mapping Hangul characters to produce an effective automated Hangul font generation system. The system was implemented using character generation engine based on deep learning CycleGAN. In order to evaluate the criteria when mapping characters in pairs, each criterion was designed based on Hangul structure and character shape, and the quality of the generated characters was evaluated. As a result of the evaluation, the standards designed based on the Hangul structure did not affect the quality of the automated Hangul font generation system. On the other hand, when tried with similar characters, the standards made based on the shape of Hangul characters produced better quality characters than when tried with less similar characters. As a result, it is better to generate automated Hangul font by designing a learning method based on mapping characters in pairs that have similar character shapes.

This study attempted to analyze human brain responses toward visual content through EEG signals and intended to measure brain wave reactions of different age groups to determine the sexuality level of the media. The experimental stimuli consist of three different video footage (rated ages 12, 15, and 18) to analyze how subjects react in situations where they actually watch sexual content. For measuring and analyzing brain wave reactions, EEG equipment records alpha, beta, and gamma wave responses of the subjects' left and right frontal lobes, temporal lobes, and occipital lobes. The subjects of this study were 28 total and they are divided into two groups. The experiment configures a sexual content classification scale with age or gender as a discriminating variable and brain region-specific response frequencies (left/right, frontal/temporal/occipital, alpha/beta/gamma waves) as independent variables. The experimental results showed the possibility of distinguishing gender and age differences. The apparent differences in brain wave response areas and bands among high school girls, high school boys, and college students are found. Using these brain wave response data, this study explored the potential of developing algorithm for measurement of age-specific responses to sexual content and apply it as a film rating.

It is that failure of equipment at the factory site causes personal injury and property damage. We are required a real-time monitoring and risk forecasting techniques to prevent for equipment failure. In this paper, we proposed a 3-phase smart plug and real-time monitoring system that can be used in factories, and collected environmental information and power information using a smart plug to analyze the data. In order to analyze the correlation between the risk situation and the collected data, we predicted the risk situation using Linear Regression, SVM, and ANN algorithms. As a result, the SVM and ANN algorithms obtained high predictive accuracy and developed a mobile app that could use it to check the risk forecast results.

In this paper, we propose a variational expectation-maximization algorithm that computes posterior probabilities from Latent Dirichlet Allocation (LDA) model. The algorithm approximates the intractable posterior distribution of a document term matrix generated from a corpus made up by 50 papers. It approximates the posterior by searching the local optima using lower bound of the true posterior distribution. Moreover, it maximizes the lower bound of the log-likelihood of the true posterior by minimizing the relative entropy of the prior and the posterior distribution known as KL-Divergence. The experimental results indicate that documents clustered to image classification and segmentation are correlated at 0.79 while those clustered to object detection and image segmentation are highly correlated at 0.96. The proposed variational inference algorithm performs efficiently and faster than Gibbs sampling at a computational time of 0.029s.