Earticle

Predicting link quality before establishing communication between transmitter and receiver enhances channel selection. With the advancements in artificial intelligence, prediction is now possible for complex environments such as riverside, maritime and polar regions. This paper evaluates Wi-Fi and LoRa radios, utilizing Received Signal Strength Indicator (RSSI) to understand link quality in riverside environments. The proposed approach compares traditional regression techniques with advanced deep learning models. Error metrics such as Root Mean Squared Error (RMSE), Mean Squared Error (MSE), and Mean Absolute Error (MAE), assess performance. The results demonstrate that ST-LSTM-CNN consistently surpasses other models for Native AI for Non-Terrestrial Networks (NTN).

본 연구에서는 지방량을 측정하는 Fat-amount 검사에서 FBP(Filtered back projection), IR(Iterative Reconstruction)과 함께 DLIR 기법을 이용하여 영상을 재구성한 후 HU(Hounsfield unit), VFA(Visceral Fat Area), SFA(Subcutaneous Fat Area), V/S(Visceral-Subcutaneous Fat Ratio), SD(Standard Deviation)을 비교하여 DLIR의 유용성을 평가했다. 지방의 정량적 측정에서는 FBP, ASIR-V와 DLIR이 통계적으로 유의한 차이 없는 결과를 도출할 수 있었고, 오히려 SD은 DLIR에서 낮아져 Fat의 정량적 측정은 동일하면서, 영상의 질을 높일 수 있는 임상적 으로 충분히 가치 있는 기법이라고 보여진다. 타 검사에 비해 큰 단점으로 지목되고 있는 방사선량을 줄이는 데에 있어 충분한 가지가 있다고 사료된다.

In this study, in ‘CT Fat-amount’ to measure the amount of fat, the image was reconstructed using the DLIR technique along with FBP (Filtered back projection) and IR (Iterative Reconstruction), and then HU (Hounsfield unit), VFA (Visceral Fat Area), and SFA The usefulness of DLIR was evaluated by comparing (Subcutaneous Fat Area), V/S (Visceral-Subcutaneous Fat Ratio), and SD (Standard Deviation). In the quantitative measurement of fat, FBP, ASIR-V, and DLIR were able to produce results with no statistically significant difference. Rather, the SD was lower in DLIR, so the quantitative measurement of fat was the same, but the quality of the image could be improved clinically. It appears to be a worthwhile technique. It is believed that there are sufficient ways to reduce the radiation dose, which is considered a major disadvantage compared to other tests.

Accurate classification of public policy is essential for effective policy analysis, design, comparison, and formulation across countries. However, manual classification by policy experts can lead to inconsistencies and human errors, highlighting the need for a more reliable and efficient approach. This study proposes a deep learning-based model to support policy classification using artificial intelligence. Leveraging Korean policy datasets, comprising administrative data (1988–2018), legislative data (1987–2018), and media data (1988–2020), previously curated by experts, we developed an AI model for automated policy classification based on the KoBERT language model. Designed as a supplementary tool for policy experts, this model enhances classification consistency, reduces decision-making time, and improves overall productivity. Moreover, the model enables the classification, comparison, and evaluation of diverse policies at both local and national levels, offering valuable support for strategic public policy development. The proposed model achieved a Top-1 accuracy of 62.4% and a Top-3 accuracy of 71.6%, outperforming traditional baselines and demonstrating its practical potential for real-world policy analysis.

최근 발생하고 있는 기후변화는 전세계적으로 많은 피해를 발생시키고 있다. 한반도의 기존 강우형태는 장시간에 걸친 강우와 발생기간을 예측할 수 있었으나, 기후 변화로 인한 강우 형태의 변화는 기존에 마련한 위기대응방안이 무용한 상황을 만들고 있다. 2022년 8월 집중호우로 인해 서울·경기·강원·충남 등 10개 지자체가 특별재난지역으로 우선 선포되었다. 도시화에 따른 불투수면적의 증가와 건물의 지하 공간의 활용 증가하고 있는 가운데 기후 변화로 인한 집중호우는 도심지 내 하수관로 설계용량을 초과하 는 경우가 빈번해지고 있으며, 범위는 국부적이지만 피해의 규모는 커지는 사례들이 증가하고 있다. 도시침수는 다양한 조건에 의해 침수범위가 결정되며 불시에 일어나는 침수의 형태로 인해 신속한 침수상황 인지 및 위기대응이 필요하다. 이를 위하여 AI기반 영상분석 기술을 통해 전국에 분포되어 있는 CCTV를 활용하여 침수상황을 상시 모니터링 할 수 있는 시스템 구축이 필요하며, 정형화된 IoT기반 실시간 계측센서 데이터와 비정형 데이터인 CCTV영상을 분석하고 연계한 새로운 형태의 도시침수 모니터링 기술이 필요하다.

Recently, climate change has caused a lot of damage worldwide. Existing rainfall patterns on the Korean Peninsula could predict long-term rainfall patterns and periods, but changes in rainfall patterns due to climate change made crisis response measures useless. Due to heavy rain in August 2022, 10 local governments, including Seoul, Gyeonggi, Gangwon, and Chungnam, were declared as special disaster areas for the first time. As the impermeable area increases due to urbanization and the use of the underground space of buildings increases, the design capacity of sewage pipes is often exceeded, and the scope is local but the damage scale is increasing. Urban floods require rapid situational awareness and crisis response. To this end, it is necessary to establish a system that can monitor the flood situation at all times using CCTV distributed across the country through AI-based image analysis technology, and a new type of urban flood monitoring technology that analyzes and links it based on standardized IoT unstructured data such as real-time measurement sensor data and CCTV images is required.

This Leishmaniasis is common skin lesion parasitic disease caused by Leishmania protozoan parasites on exposed body and its polymorphic nature complicates to diagnosis because the lesion may create confusion with other dermatoses likewise fungi, bacteria and non-infectious diseases. The molecular techniques, microscopy, culture, and rapid diagnostic test are conventional methods that are timeconsuming, expensive, susceptible to errors with limited resources in health care services. Early diagnosis with timely identification of multifaceted Leishmaniasis is aided to selection of therapy and provide comfort to patient to combating with it. The promising integration of artificial intelligence (AI) with medical diagnostics has efficacy in numerous fields of identification of diseases as in Dermatology research. The fast, efficient and automatic diagnosing of leishmaniasis with microscopic images of lesion's seamer with VGG-16 deep learning (DL) model is the approach to reach the objective of this designed research to identify the negative and positive results. The exceptional performance of designed VGG-16 is achieved with accuracy of 88.14%, precision 100%, sensitivity 77.42%, specificity 100%, F1-score 0.87%, and ROC curve 97%. The proposed modified VGG-16 model is more precise, swift, reliable, efficient, effectual, economical and user-friendly substitute to address all key factors than human resource to find the leishmaniasis affected that may support medical care services.

연구목적: 본 연구는 발전소에서 반복 발생하는 중대재해(SIF: 떨어짐, 끼임, 부딪힘) 예방을 위해 딥러 닝 기반 AI CCTV 관제시스템을 구축하고 효과를 실증적으로 검증하는데 있다. 연구방법: 2018~2022 년 안전사고 데이터와 2023~2025년 탐지율 성능데이터를 활용하여 기술통계, 추세분석, 이중차분 (DiD), 이벤트 스터디 모형으로 AI도입 전후 효과를 분석한다. 연구결과: 중대재해는 전체 사고의 46.8%를 차지하였으며, 도입부서의 안전모 탐지율은 23.78%→97.56%, 쓰러짐은 37.79%→87.79%로 향상되었다. DiD 분석결과, AI 도입 효과( +0.368(p<0.01)가 확인되었고, 이벤트 스터디 분석에서도 효 과가 점차 강화되고 내재화됨을 입증하였다. 결론: 안전사고 통계와 AI 성능데이터 결합을 통해 사고예 방효과를 계량적으로 입증하였고, AI CCTV 감시체계는 인력중심 감시한계를 극복하고 안전관리를 사 후 대응에서 사전 예방중심으로 전환해야함을 보여준다. 이러한 스마트 안전관리체계는 고위험 현장 의 중대재해 예방에 기여할 것이다.

Purpose: This study aims to empirically verify the effectiveness of a deep learning–based AI CCTV control system in preventing recurring serious industrial accidents(SIFs: falls, entrapments, and collisions) at power plants. Method: Industrial accident data from 2018–2022 and AI CCTV detection performance data from 2023–2025 were analyzed. Descriptive statistics, trend analysis, Difference-in- Differences(DiD), and event study models were applied to assess pre- and post-adoption effects. Result: SIFs accounted for 46.8% of all industrial accidents. After AI CCTV implementation, helmet detection improved from 23.78% to 97.56%, and fall detection from 37.79% to 87.79%. The DiD analysis showed a significant adoption effect of +0.368(p<0.01). Event study analysis further confirmed that the preventive effects gradually strengthened and became institutionalized. Conclusion: By integrating accident statistics with AI performance data, this study quantitatively demonstrated the accident prevention effect of AI CCTV. The system overcame the limitations of human-centered monitoring and shifted safety management from reactive response to proactive prevention. This smart safety management system will help prevent serious accidents in high-risk sites.

친환경 연료의 필요성이 대두되면서 화석연료의 사용이 줄어들고 있다. 이에 따라 해양경찰은 국가기조에 맞게 최근 해양오염 방제선을 LNG 선박으로 건조하는 등 연료 시스템 전환에 힘쓰고 있다. 그러나 기존연료대비 암모니아, 메탄올, 수소, LNG와 같은 인화성 혹은 독성이 높은 연료를 선박 연료로 사용할 경우, 누출 사고 에 대비한 방안을 마련해야 한다. 이에 따라 세계적으로 친환경 연료 사용과 첨단 해양 모빌리티 기술 개발 및 안전에 관한 연구가 활발히 진행되고 있다. 따라서 기 존 MDO 연료를 사용하는 실습선의 연료를 암모니아 연료로 가정하여 피해범위를 예측하였다. 이와 더불어 AI 딥러닝 기반의 전처리 시스템 모듈을 구축 및 경로 탐 색 알고리즘을 활용해 인원수, 통로폭, 선체 기울기에 따른 최적 대피 경로를 도출하 였다. 가스 누출에 따른 피난 시, 딥러닝을 통해 구역별 최대 인구밀도가 5.4 per/㎡ 에서 일반적인 경로 탐색 방식과 비교한 결과 2.6 per/㎡로 개선 됨을 확인하였다. 이에 따라 전체 피난 시간이 1,030초에서 248초로 감소되었으며 기존 대비 예상 사망자 수가 67명 줄어 생존율이 35% 향상되었다. 따라서 본 연구에서 제안한 프로그램은 선박재난 상황에서 최적의 대피 경로를 도출하여 승객 및 승조원의 안전을 확보하는 의사결정 지원 기초 자료로 활용될 수 있을 것으로 보인다.

As the need for eco-friendly fuel arises, the use of fossil fuels is decreasing. Accordingly, the korea coast guard are striving to convert the fuel system by building the marine pollution control ship with LNG ships. However, when highly flammable fuels such as LNG and hydrogen are used as ship fuel, it is necessary to prepare a plan for a leakage accident. Against this background, research on the use of eco-friendly fuels and the development and safety of advanced marine mobility technologies is being actively conducted. Therefore, in this study, a post-processing system module based on a convolutional neural network was constructed, and the optimal evacuation route according to the number of people, passage width, and hull trim was derived using a deep learning model and path search algorithm based on a deep neural network. As a result of comparing it with the classical pathfinding method under the assumed conditions, the maximum population density by area was improved from 5.4 per/㎡ to 2.6 per/㎡ through deep learning. In addition, the total evacuation time decreased from 1,030 seconds to 248 seconds, and the estimated number of deaths decreased by 67, improving the survival rate by 35%. Therefore, the program developed in this study is expected to be used as decision support data to secure the safety of passengers and crew by deriving the optimal evacuation route in a disaster situation.

Injection-molded products frequently exhibit localized surface defects such as weld lines, flow marks, scratches, bubbles, and burn marks due to variations in material flow, mold temperature, and cooling conditions. Conventional visual inspection is highly dependent on operator experience, while rule-based machine vision methods are limited under variations in lighting and surface texture. This study proposes a deep learning–based defect detection model using YOLOv8 combined with a novel Defect-Aware Augmentation technique designed to enhance robustness for small, local defect regions. The proposed augmentation pipeline includes geometric transformations, optical perturbations, local defect patch synthesis, and diffusion-based synthetic defect generation. Experiments were conducted on a custom dataset of 5,000 images (3,000 normal and 2,000 defective). Results show that the proposed model achieves significant improvements over baseline models, obtaining 95% precision, 90% recall, and 0.96 mAP@0.5, outperforming the default YOLOv8 model by 7%p in mAP. Ablation studies verify that defect-aware augmentation is the dominant factor contributing to the performance gain. The proposed system demonstrates high applicability for automated quality inspection in injection- molding production lines.

본 연구는 초등학생의 딥러닝 개념 학습을 위한 교육 프로그램을 개발하는 것이다. 교육 프로그램의 모델은 CT요소 중심 모델을 토대로 딥러닝 교수학습모델을 개발하였다. 개발한 프로그램의 주제는 인공지능의 이미지 인식 CNN알고리즘으로 정하고, 9개 차시 교육프로그램을 개발하였다. 프로그램은 6학년을 대상으로 2주간에 걸 쳐 적용을 하였다. 프로그램에 대한 학습 적합도 검사는 전문가 타당도 분석 결과로 CVR이 타당하게 나왔다. 학습자 수준 적합도와 교사 지도 수준의 적합도 문항의 경우 .80이하로 나타났으며 .96이 넘은 학습 환경과 매체 의 적합도 문항에서는 높게 나타났다. 학생들의 만족도 분석 결과 학습의 이해도와 유익성, 흥미도, 학습자료 등에 대해서 평균 4.0이상을 보여 긍정적인 평가를 하여 본 연구의 가치를 확인할 수 있었다.

The purpose of this study is to develop an educational program for learning deep learning concepts for elementary school students. The model of education program was developed the deep-learning teaching method based on CT element-oriented teaching and learning model. The subject of the developed program is the artificial intelligence image recognition CNN algorithm, and we have developed 9 educational programs. We applied the program over two weeks to sixth graders. Expert validity analysis showed that the minimum CVR value was more than .56. The fitness level of learner level and the level of teacher guidance were less than .80, and the fitness of learning environment and media above .96 was high. The students' satisfaction analysis showed that students gave a positive evaluation of the average of 4.0 or higher on the understanding, benefit, interest, and learning materials of artificial intelligence learning.

Always, emotion is mutant. This is principle of literary treatment. In the literature, sadness is not sadness, and ‘loving emotion’ is not ‘loving emotion.’ Despite loving of our, loving is sadness. Also loving is to cry. This crying becomes love. This study is to show the mutant emotion which is to be able to code Deep Learning AI. We explored the Sijo “Streams that cried last night", because this Sijo was useful to study mutant emotion. The result was that this Sijo was coding the mutant emotion. Almost continuously, the sadness codes were spawning and concentrating. So this Sijo was making the emotion of love with the sadness. If this study is continued, It is believed that our lives will be much happier. And the method of literary therapy will be able to more upgrade.

본 논문에서는 딥러닝 기반의 개인 맞춤형 실버세대 케어 서비스 애플리케이션을 설계하고 구현한다. 이 애플리케이션은 사용자의 편의성을 고려하여 STT(Speech to Text) 기술을 사용해 사용자의 발화를 텍스트로 변환하고, 이를 Microsoft 사의 대화형 멀티 에이전트 거대 언어 모델인 Autogen의 입력으로 사용한다. Autogen은 사용자와 ChatBot의 대화 데이터를 활용하여 상대방의 의도를 파악하여 답변에 대하여 응답한다. 그리고 백엔드 에이전트를 활용하여 위시리스트, 공유 달력 그리고 보이스 클로닝을 위한 딥러닝 모델을 통해 상대방의 목소리가 담긴 안부 메시지 기능을 제공한다. 또한, 애플리케이션은 SKT 사의 인공지능 누구(NUGU) 스피커를 탑재하여 홈 IoT 서비스 기능을 제공한다. 이러한 기능을 통해 제안하는 지능형 애플리케이션은 향후 미래 인공지능 기반의 실버세대 케어 기술에 기여할 것이다.

In this paper, we propose a deep learning-based personalized senior care service application. The proposed application uses Speech to Text technology to convert the user's speech into text and uses it as input to Autogen, an interactive multi-agent large-scale language model developed by Microsoft, for user convenience. Autogen uses data from previous conversations between the senior and ChatBot to understand the other user's intent and respond to the response, and then uses a back-end agent to create a wish list, a shared calendar, and a greeting message with the other user's voice through a deep learning model for voice cloning. Additionally, the application can perform home IoT services with SKT's AI speaker (NUGU). The proposed application is expected to contribute to future AI-based senior care technology.

Soil texture is determined by the proportions of sand, silt, and clay within the soil, which influence characteristics such as porosity, water retention capacity, electrical conductivity (EC), and pH. Traditional classification of soil texture requires significant sample preparation including oven drying to remove organic matter and moisture, a process that is both time-consuming and costly. This study aims to explore an alternative method by developing an AI model capable of predicting soil texture from images of pre-sorted soil samples using computer vision and deep learning technologies. Soil samples collected from agricultural fields were pre-processed using sieve analysis and the images of each sample were acquired in a controlled studio environment using a smartphone camera. Color distribution ratios based on RGB values of the images were analyzed using the OpenCV library in Python. A convolutional neural network (CNN) model, built on PyTorch, was enhanced using Digital Image Processing (DIP) techniques and then trained across nine distinct conditions to evaluate its robustness and accuracy. The model has achieved an accuracy of over 80% in classifying the images of pre-sorted soil samples, as validated by the components of the confusion matrix and measurements of the F1 score, demonstrating its potential to replace traditional experimental methods for soil texture classification. By utilizing an easily accessible tool, significant time and cost savings can be expected compared to traditional methods.

Purpose: The purpose of this study is to actually implement and verify whether welding defects can be detected in real time by utilizing deep learning AI solutions in the welding process of electric vehicle hairpin winding motors. Methods: AI's function and technological elements using synthetic neural network were applied to existing electric vehicle hairpin winding motor laser welding process by making special hardware for detecting electric vehicle hairpin motor laser welding defect. Results: As a result of the test applied to the welding process of the electric vehicle hairpin winding motor, it was confirmed that defects in the welding part were detected in real time. The accuracy of detection of welds was achieved at 0.99 based on mAP@95, and the accuracy of detection of defective parts was 1.18 based on FB-Score 1.5, which fell short of the target, so it will be supplemented by introducing additional lighting and camera settings and enhancement techniques in the future. Conclusion: This study is significant in that it improves the welding quality of hairpin winding motors of electric vehicles by applying domestic artificial intelligence solutions to laser welding operations of hairpin winding motors of electric vehicles. Defects of a manufacturing line can be corrected immediately through automatic welding inspection after laser welding of an electric vehicle hairpin winding motor, thus reducing waste throughput caused by welding failure in the final stage, reducing input costs and increasing product production.

본 연구는 2015년부터 2024년까지 발표된 인공지능(AI) 분야 학술 논문을 대상으로 인용 네트워크 분석(citation network analysis)을 실시하여 연구 주제의 진화, 핵심 연구자의 영향력, 협력 네트워크 구조를 실증적으로 탐색하였다. 분석 자료는 Dimensions.ai 데이터베이스로부터 수집된 6,000여 편의 논문이며, VOSviewer 소프트웨어를 활용해 중심성 분석, 클러스터링, 네트워크 시각화, 그리고 통계적 상관·회귀분석을 수행하였다. 연구 결과 AI 연구는 초기 딥러닝 중심에서 설명 가능한 AI(XAI), 인지 기반 AI, 대형 언어 모델(LLM) 중심으로 주제가 확장되었으며, 인간 중심성과 해석 가능성을 강조한 논문들이 높은 중심성을 보였다. 협업 구조 분석에서는 소규모 독립 연구에서 다자간 공동연구 중심으로의 전환이 확인되었고, 네트워크 지표 간에는 유의미한 양의 상관관계(r=0.19, p<.01)가 나타났으며, 회귀분석에서도 총 링크 강도가 정규화 피인용 수를 유의하게 예측하는 경향이 확인되었다. 이러한 결과는 키워드 빈도에 기반한 단편적 분석을 넘어 인용 구조를 활용한 정량·정성 통합 분석이 AI 연구 생태계의 구조적 특성과 진화 방향을 보다 입체적으로 조망할 수 있음을 보여준다. 본 연구는 향후 AI 연구 정책 수립, 융합 연구 촉진, 그리고 학제 간 협력 전략 설계에 실증적 근거를 제공한다.

This study conducts a citation network analysis of academic publications in the field of AI from 2015 to 2024, examining the evolution of research topics, the influence of key researchers, and the structure of collaboration networks. The dataset comprises over 6,000 papers retrieved from the Dimensions.ai database, analyzed using VOSviewer software to perform centrality analyses, clustering, network visualization, and statistical correlation and regression analyses. Results show that AI research has evolved from an early focus on deep-learning to topics such as explainable AI (XAI), cognitive-based AI, and large language models (LLMs), with papers emphasizing human-centeredness and interpretability exhibiting higher centrality. Collaboration network analysis revealed a shift from small-scale independent studies to multi-author collaborative research, while statistical analysis identified a significant positive correlation between total link strength and normalized citation counts(r=0.19, p<.01), with regression analysis confirming link strength as a significant predictor of citation impact. Moving beyond keyword-based frequency analysis, this study applies a structural, citation-based approach to provide empirical evidence for AI research policy, foster interdisciplinary collaboration, and guide strategic planning for sustainable research ecosystems.

최근 건조한 날씨에 생기는 불씨가 초기진압에 실패하며 대형 산불로 확산되는 사고가 많이 발생하고 있다. 산에서 생기는 불씨를 초반에 발견하는 것으로 문제 해결이 가능해진다. 넓은 지역을 사람이 직접으로 관찰하는 것은 어렵기 때문에 자동 검출 시스템이 필요하다. 영상 분석을 통한 자동 산불 감지 시스템을 제안하고자 한다. 이를 통해 인력 낭비를 줄일 것으로 예상하며 기존 방식의 한계점을 개선한 학습법이 효과가 있음을 검증한다. 감지 성능 평가 지표로 객체 탐지 예측의 정확률, 재현율과 mAP에서 높은 성과를 달성하였다. 그리고 발화 연기와 아닌 것을 구별하는 학습하는 방식을 제안한다. 분할기법을 이용한 본 논문의 예측방법은 면적을 계산하여 산불 규모를 유추할 수 있을 것으로 기대된다.

Recently, a lot of accidents have occurred in which small embers generated in dry weather fail to suppress the initial fire and spread to large wildfires. To solve this problem, finding embers in the mountains at the beginning is important. An automatic detection system is needed because it is difficult for humans to observe large areas directly. This study was conducted to find a highly practical method by proposing an AI-automatic forest wildfire detection system using image analysis. Through this, it is expected to reduce manpower waste drastically, and it is verified that the learning method that improves the limitations of the existing method has a real effect. High performance was achieved in the accuracy rate, recall rate, and mean average precision of object detection prediction, which can be seen as a wildfire detection performance evaluation index. This paper proposes an AI model(YOLO, You Only Looks Once model) method of deep learning to distinguish between ignition smoke and non-ignition smoke. In the future, it is expected that the scale of forest fire can be inferred by calculating the distance between the forest fire occurrence point and the shooting point and the area using the segmentation technique installed in the AI model.

This paper explores the integration of artificial intelligence and causal inference in social science research, focusing on causal deep learning. We examine key theories including Pearl's Structural Causal Model, Rubin's Potential Outcomes Framework, and Scholkopf's Causal Representation Learning. Methodologies such as structural causal models with deep learning, counterfactual reasoning, and causal discovery algorithms are discussed. The paper presents applications in social media analysis, economic policy, public health, and education, demonstrating how causal deep learning enables nuanced understanding of complex social phenomena. Key challenges addressed include model complexity, causal identification, interpretability, and ethical considerations like fairness and privacy. Future research directions include developing new AI architectures, real-time causal inference, and multi-domain generalization. While limitations exist, causal deep learning shows significant potential for enhancing social science research and informing evidence-based policy-making, contributing to addressing complex social challenges globally.

The rapid advancement of Artificial Intelligence(AI) has achieved remarkable progress in various fields such as image editing, audio generation, and video manipulation. However, it has also introduced new security threats, including deepfake speech and voice spoofing. This paper proposes a multi-feature deep learning based AI synthetic speech detection method capable of addressing these threats with high accuracy. The ASVspoof 2021 dataset's Logical Access(LA) and Deepfake(DF) data were used for training and testing. The proposed system utilizes two audio features, Mel-Spectrogram and MFCC(Mel-Frequency Cepstral Coefficients), to convert audio data into visual and sequential forms for training and inference. To demonstrate the superiority of the proposed method, a comparative analysis was conducted with various models such as CNN, BiLSTM, Transformer, and ensemble methods. Experimental results showed that the multi-feature fusion model outperformed single and ensemble models. The proposed multi-feature fusion model, which combines ConvNeXt-base and BiLSTM using a Late Fusion approach, achieved the highest performance with an accuracy of 98.44 %. The method proposed in this paper is expected to serve as a key technology in future AI deepfake synthetic speech detection systems.

目前，以深度學習為代表的高新技術快速崛起，人類社會正在進入人工智能時代。文創產品的圖案設計可以借助深度學習進行發展，這將改善現有的不足，促進相關行業升級。本研究首先梳理了深度學習中能夠進行結合應用的技術。然後將新技術融入現有的設計流程，形成了基於深度學習的文創產品圖案設計框架。為驗證所提出框架的有效性，召集設計從業人員進行了對照實驗。實驗結果表明，本研究所提出的框架有效利用了人工智能的技術優勢，能夠提高設計作品的創意度和綜合效果，能夠縮短設計所需時間，提高設計師的設計效率。同時，所提出的框架降低了對設計師基礎能力的依賴，使設計專業的初學者在更短的時間內設計出更高質量的作品。在未來的研究中，可以將此框架推廣到實際設計工作中，並調查設計師的使用感受。

In recent years, with the rapid rise of high and new technology represented by deep learning (DL), human society is entering the era of artificial intelligence (AI). The graphic design of Cultural and Creative Products can be developed with the help of DL, which will improve the existing shortcomings and promote the upgrading of relevant industries. Firstly, this study combs the AI technologies that can be applied to design. Then the technologies are integrated into the existing design process to form a DL based framework. In order to verify the effectiveness of the proposed framework, design practitioners were called to conduct a control experiment. The results show that the proposed framework makes effective use of the technical advantages of AI, and can improve the creativity and comprehensive effect, shorten the design time, and improve the design efficiency of designers. At the same time, the proposed framework reduces the dependence on the basic ability of designers, thus beginners majoring in design can design higher quality works in a shorter time. In the future research, the proposed framework will be extended to the actual design work, and the use feelings of designers can be investigated.

인천공항의 출입국 관리 시스템 개발과 검증에 생체정보와 딥러닝 알고리즘을 활용하겠다는 정부의 계획은 한국 사회의 개인정보보호와 인공지능 정책에 대한 까다로운 질문을 우리에게 던졌다. 이 논문은 법무부 출입국․외국인정책본부와 과학기술정보통신부, 행정안전부, 개인정보보호위원회 등이 지향하는 원칙이 충돌하고 있다고 주장한다. 또 이 논문은 지능정보사회를 추진하고 디지털 플랫폼 정부를 앞당기기 위한 공공데이터 활용 정책이 생체정보를 이용한 감시를 견제하고, 개인정보보호와 정보의 자기 결정권을 보장하는 개인정보보호 정책과 잠재적 갈등 관계에 처했음을 보일 것이다. 이른바 인공지능 시대를 맞이하여 국경 앞에서 중단되지 않는 민주주의 원칙과 차별을 줄이기 위한 인권 정책에 부합하는 공공데이터 정책의 가능성을 찾기 위한 방법을 모색해야 할 시점이다.

The controversy on the border control system based on biometric information and deep learning in Incheon airport raised thorny questions on conflicts between public data utilization. This paper first lays out the detailed issues against the Korean discursive landscape of privacy and emerging technology policies. Then I will argue that the Ministry of Justice’s Korea Immigration Service, the Ministry of Science and ICT, and the Ministry of the Interior and Safety formulated and maintained conflicting policy principles. This paper also posits that the potential conflict exists between the policy for public data utilization toward intelligent information society and digital platform government on the one hand and the personal information protection policy that guarantees the right to informational self-determination on the other hand. This paper ends with an inquiry into possibilities of public data policy that resonates with the democratic principle that does not stop at the border and human rights policies for less discrimination.