Agriculture is being damaged by harmful animals such as wild boars and water deer. It need to get permission to catch a wild boar and farmers are using a lot of methods to chase harmful animals. The methods through deep learning and image processing capture harmful animals with cameras. It is difficult to analyze harmful animals that are active at night. In this case, In this case, using deep learning by image correction can achieve a higher recognition rate.

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.

As part of the cell division method, we proposed a method for segmenting images generated by topography microscopes through deep learning-based feature generation and graph segmentation. Hybrid vector shapes preserve the overall shape and boundary information of cells, so most cell shapes can be captured without any post-processing burden. NIH-3T3 and Hela-S3 cells have satisfactory results in cell description preservation. Compared to other deep learning methods, the proposed cell image segmentation method does not require postprocessing. It is also effective in preserving the overall morphology of cells and has shown better results in terms of cell boundary preservation.

Attention Deficit Hyperactivity Disorder (ADHD) is one of the most common neurological disorders in children. The diagnosis of ADHD in children is based on the interviews and observation reports of parents or teachers who have stayed with them. Since this approach cannot avoid long observation time and the bias of observers, another approach based on Electroencephalography(EEG) is emerging. The goal of this study is to develop an assistive tool for diagnosing ADHD by EEG classification. This study explores the frequency bands of EEG and extracts the implied features in them by using the proposed CNN. The CNN architecture has three Convolution-MaxPooling blocks and two fully connected layers. As a result of the experiment, the 30-60 Hz gamma band showed dominant characteristics in identifying EEG, and when other frequency bands were added to the gamma band, the EEG classification performance was improved. They also show that the proposed CNN is effective in detecting ADHD in children.

The drone technology, which is receiving a lot of attention due to the 4th industrial revolution, requires an Unmanned Aerial Vehicles'(UAVs) flight path search algorithm for automatic operation and driver assistance. Various studies related to flight path prediction and recommendation algorithms are being actively conducted, and many studies using the A-Star algorithm are typically performed. In this paper, we propose an Optimal 3D Flight Path Recommendation System for unmanned aerial vehicles. The proposed system was implemented and simulated in Unity 3D, and by indicating the meaning of the route using three different colors, such as planned route, the recommended route, and the current route were compared each other. And obstacle response experiments were conducted to cope with bad weather. It is expected that the proposed system will provide an improved user experience compared to the existing system through accurate and real-time adaptive path prediction in a 3D mixed reality environment.

A research on agricultural automation is a main issues to overcome the shortage of labor in Korea. A sweet persimmon farmers need much time and labors for classifying profitable sweet persimmon and ill profitable products. In this paper, we propose a mixed two-step synthetic neural network model for efficiently classifying sweet persimmon images. In this model, we suggested a surface direction classification model and a quality screening model which constructed from image data sets. Also we studied Class Activation Mapping(CAM) for visualization to easily inspect the quality of the classified products. The proposed mixed two-step model showed high performance compared to the simple binary classification model and the multi-class classification model, and it was possible to easily identify the weak parts of the classification in a dataset.

The post-corona era is having a major impact on university campus life, and campus cultural life has undergone drastic changes. In the process of adapting to this change. The existing face-to-face learning method is converted to non-face-to-face learning, and kiosks representing campus self-service are increasingly used. Subsequently, in order to match the characteristics of public places such as university campuses and provide better self-service, there are no related studies considering UI design elements of kiosks and characteristics of kiosks. This study conducted a survey on Chinese university students considering the characteristics and usability of university campus kiosks through user interface design elements in terms of design by considering previous studies. The purpose of this study is to examine the relationship between color, icon, typography, which are elements of kiosk interface design, and convenience, reliability, and usability, which are characteristics of a kiosk. In this study, the design elements and characteristics of the unmanned self-service system kiosk UI used by students at four universities in Korea and China were set as the scope of the study, and related documents were organized and surveyed in Korea and China. Case analysis and comparison method Through the kiosk usage characteristics of 4 universities were derived. The usability evaluation criteria use the five evaluation criteria suggested by the authority JacobNidsn. The usability evaluation criteria use the five evaluation criteria suggested by the authority JacobNidsn. The survey and analysis of this study were conducted for students who had experience using kiosks at two universities in China, and kiosk UI design elements were organized through survey and analysis of a total of 200 people.

The video face recognition (FR) is one of the most popular researches in the field of computer vision due to a variety of applications. In particular, research using the attention mechanism is being actively conducted. In video face recognition, attention represents where to focus on by using the input value of the whole or a specific region, or which frame to focus on when there are many frames. In this paper, we propose a novel attention based deep learning method. Main novelties of our method are (1) the use of combining two loss functions, namely weighted Softmax loss function and a Triplet loss function and (2) the feasibility of end-to-end learning which includes the feature embedding network and attention weight computation. The feature embedding network has a positive effect on the attention weight computation by using combined loss function and end-to-end learning. To demonstrate the effectiveness of our proposed method, extensive and comparative experiments have been carried out to evaluate our method on IJB-A dataset with their standard evaluation protocols. Our proposed method represented better or comparable recognition rate compared to other state-of-the-art video FR methods.

This paper focuses use context-sensitive spelling error correction using generative adversarial network. Generative adversarial network[1] are attracting attention as they solve data generation problems that have been a challenge in the field of deep learning. In this paper, sentences are generated using word embedding information and reflected in word distribution representation. We experiment with DCGAN[2] used for the stability of learning in the existing image processing and D2GAN[3] with double discriminator. In this paper, we experimented with how the composition of generative adversarial networks and the change of learning corpus influence the context-sensitive spelling error correction In the experiment, we correction the generated word embedding information and compare the performance with the actual word embedding information.

In this paper, we investigate automated data validation techniques for artificial intelligence training, and also disclose open-source frameworks, such as Google's TensorFlow Data Validation (TFDV), that support automated data validation in the AI model development process. We also introduce an experimental study using public data sets to demonstrate the effectiveness of the open-source data validation framework. In particular, we presents experimental results of the data validation functions for schema testing and discuss the limitations of the current open-source frameworks for semantic data. Last, we introduce the latest studies for the semantic data validation using machine learning techniques.

In this paper, the importance of input factors of a DNN (Deep Neural Network) PM_2.5 forecasting model using LRP(Layer-wise Relevance Propagation) is analyzed, and forecasting performance is improved. Input factor importance analysis is performed by dividing the learning data into time and PM_2.5 concentration. As a result, in the low concentration patterns, the importance of weather factors such as temperature, atmospheric pressure, and solar radiation is high, and in the high concentration patterns, the importance of air quality factors such as PM_2.5, CO, and NO₂ is high. As a result of analysis by time, the importance of the measurement factors is high in the case of the forecast for the day, and the importance of the forecast factors increases in the forecast for tomorrow and the day after tomorrow. In addition, date, temperature, humidity, and atmospheric pressure all show high importance regardless of time and concentration. Based on the importance of these factors, the LRP_DNN prediction model is developed. As a result, the ACC(accuracy) and POD(probability of detection) are improved by up to 5%, and the FAR(false alarm rate) is improved by up to 9% compared to the previous DNN model.

Lately, the mental issues of Chinese worldwide understudies have become an unavoidable examination issue. As per important explores in different nations, the level of discouragement inclination of Chinese understudies has been on the ascent. This examination utilizes the Bread cook Wretchedness Scale II to test the psychological condition of 249 Chinese understudies concentrating in South Korea. The outcomes show that Chinese understudies in South Korea likewise have a specific level of mental issues. The extent has reached 34.5% of the absolute number of Chinese understudies in South Korea. This exploration investigates the current circumstance dependent on the review results, and directs a top to bottom examination of the attributes of worldwide understudies in Korea. Simultaneously, it is recommended that mental treatment can be utilized to work on their psychological state. Through the pertinent inquiries for grounds mental directing administrations, it obviously presents the issues that Chinese global understudies face in South Korean grounds mental guiding, just as the school's improvement measures. Simultaneously, this examination consolidates the media SNS social stage, investigates existing issues, and proposes how schools can adequately construct a SNS stage at the degree of mental advising, just as the adapting techniques and systems for Chinese understudies dealing with mental issues.

Recent years, due to the direct or indirect damages caused by meteorological disasters more and more attention have been paid to natural disasters. At same time, diversified and multi-sensory interactive meteorological services is increasingly demanded. In this study, novel interactive meteorological service was compared with the traditional communication methods. Combining with case studies and systems creation, a virtual reality weather simulation framework was proposed, and a realistic virtual game environment providing real-time and historical weather information was created. The primary goal of this study is to build a weather display cabinet game system by using virtual reality technology, and promoting public's understanding of the principles of weather changes. With the interactive games in realistic scenarios, public's awareness for disasters prevention could be promoted. It is helping to change public's traditional understanding of meteorological theories, and will provide a more convenient way for the public to explore more effective weather forecasts. The simulation system is supported by VR technology. It was combined with Leap Motion interactive equipment to make popularization games for weather science. T-test data analysis showed that the application of VR technology in weather games has strong operability and interactivity.

The visual design elements of packaging play an vital role in attracting consumers' attention, forming their sense of pleasure and sensory expectations, as well as having an important impact on the actual product experience. This study aims to explore whether the colors and textures in the packaging design elements of canned beer will affect consumers' senses, judgments and taste perception of canned beer. Participants in Survey I (N=193) evaluated their expectations for beer freshness, softness, bitterness and alcohol content by observing the packaging of canned beer, then recorded whether they could get the expected experience when tasting beer, and evaluated their preference and purchase intention. In the Survey II, 4 kind of colors and 2 types of texture shapes were applied, and all participants were provided with the same beer for multivariate cross contract analysis to evaluate the specific impact of color and texture of packaging design elements on consumers' taste and the interaction between sensory judgment and taste perception.The results showed that both colors and textures had a significant affect on the sensory expectation (pre-tasting score) of canned beer, but their effects on the sensory score after tasting (actual perception)were inobvious. The analysis of the influence of the interaction between color, texture and shape on taste perception shows that when the expected packaging appearance perception is similar to the actual drinking perception (i.e., straight line and B/G or arc line and Y/R), it is more likely to obtain consumers' favor and higher purchase intention evaluation than the inconsistent canned beer packaging (i.e., straight line and Y/R and arc line and B/G). This paper discusses the influence of these results on the packaging design of canned beer and the possibility of improving the brand efficiency by meeting the visual elements of packaging design expected by consumers.