• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.851-853
• /
• 2022

고령화로 인해 65세 이상 운전자가 급증하며 고령운전자의 교통사고 비율이 증가함에 따라 시급한 사회 문제로 떠오르고 있다. 이에 본 연구에서는 객체 검출, 인식 모델을 결합하고 신호등을 인식하여 Text-To-Speech(TTS)로 알리는 쿠버네티스 기반의 프레임워크를 제안한다. 객체 검출 단계에서는 YOLOv5 모델들의 성능을 비교하여 활용하였으며 객체 인식 단계에서는 C-RNN 기반의 attention-OCR 모델을 활용하였다. 이는 신호등의 내부 LED 영역이 아닌 이미지 전체를 인식하는 방식으로 오탐지 요소를 낮춰 인식률을 높였다. 결과적으로 1,628장의 테스트 데이터에서 accuracy 0.997, F1-score 0.991의 성능 평가를 얻어 제안한 프레임워크의 타당성을 입증하였다. 본 연구는 후속 연구에서 특정 도메인에 딥러닝 모델을 한정하지 않고 다양한 분야의 모델을 접목할 수 있도록 하며 고령 운전자 및 신호 위반으로 인한 교통사고 문제를 예방할 수 있다.

• Korean Journal of Artificial Intelligence
• /
• v.11 no.4
• /
• pp.29-34
• /
• 2023

The flow from developing a machine learning model to deploying it in a production environment suffers challenges. Efficient and reliable deployment is critical for realizing the true value of machine learning models. Bridging this gap between development and publication has become a pivotal concern in the machine learning community. FastAPI, a modern and fast web framework for building APIs with Python, has gained substantial popularity for its speed, ease of use, and asynchronous capabilities. This paper focused on leveraging FastAPI for deploying machine learning models, addressing the potentials associated with integration, scalability, and performance in a production setting. In this work, we explored the seamless integration of machine learning models into FastAPI applications, enabling real-time predictions and showing a possibility of scaling up for a more diverse range of use cases. We discussed the intricacies of integrating popular machine learning frameworks with FastAPI, ensuring smooth interactions between data processing, model inference, and API responses. This study focused on elucidating the integration of machine learning models into production environments using FastAPI, exploring its capabilities, features, and best practices. We delved into the potential of FastAPI in providing a robust and efficient solution for deploying machine learning systems, handling real-time predictions, managing input/output data, and ensuring optimal performance and reliability.

• The Journal of Information Systems
• /
• v.33 no.1
• /
• pp.91-114
• /
• 2024

Purpose The purpose of this study is to empirically investigate the factors that influence users' continuous intention to use ChatGPT based on the Expectation Confirmation Model(ECM). Drawing from the literature, this study identifies anthropomorphism and trust as key characteristics of generative AI and ChatGPT. Design/methodology/approach The research model was developed based on ECM and literature research to investigate the impacts of anthropomorphism and trust on continuous intention of using ChatGPT. In order to test the hypothese, a total of 193 questionnaires were collected and analyzed for the structural equation modeling with SmartPLS 4.0. Findings The study's findings show that all proposed hypotheses were supported, suggesting that the ECM is a valid framework for examining continuous intention of using ChatGPT. Moreover, the study stressed the crucial role of anthropomorphism in the model, showing the positive impact on expectation confirmation, perceived usefulness, and trust in ChatGPT. Also, trust positively affects perceived usefulness. These findings provide valuable insights for enhancing user satisfaction and continuous usage intention, serving as a foundation for development strategies for ChatGPT and similar AI-based systems.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.23 no.3
• /
• pp.236-251
• /
• 2020

As remote sensing technologies are evolving, and more satellites are orbited, the demand for using satellite data for disaster monitoring is rapidly increasing. Although natural and social disasters have been monitored using satellite data, constraints on establishing an integrated satellite-based near real-time disaster monitoring system have not been identified yet, and thus a novel framework for establishing such system remains to be presented. This research identifies constraints on establishing satellite data-based near real-time disaster monitoring systems by devising and testing a new conceptual framework of disaster monitoring, and then presents a feasible disaster monitoring system that relies mainly on acquirable satellite data. Implementing near real-time disaster monitoring by satellite remote sensing is constrained by technological and economic factors, and more significantly, it is also limited by interactions between organisations and policy that hamper timely acquiring appropriate satellite data for the purpose, and institutional factors that are related to satellite data analyses. Such constraints could be eased by employing an integrated computing platform, such as Amazon Web Services(AWS), which enables obtaining, storing and analysing satellite data, and by developing a toolkit by which appropriate satellites'sensors that are required for monitoring specific types of disaster, and their orbits, can be analysed. It is anticipated that the findings of this research could be used as meaningful reference when trying to establishing a satellite-based near real-time disaster monitoring system in any country.

• Journal of Korea Multimedia Society
• /
• v.24 no.8
• /
• pp.979-987
• /
• 2021

Multiple Sclerosis (MS) can be early diagnosed by detecting lesions in brain magnetic resonance images (MRI). Unsupervised anomaly detection methods based on autoencoder have been recently proposed for automated detection of MS lesions. However, these autoencoder-based methods were developed only for 2D images (e.g. 2D cross-sectional slices) of MRI, so do not utilize the full 3D information of MRI. In this paper, therefore, we propose a novel 3D autoencoder-based framework for detection of the lesion volume of MS in MRI. We first define a 3D convolutional neural network (CNN) for full MRI volumes, and build each encoder and decoder layer of the 3D autoencoder based on 3D CNN. We also add a skip connection between the encoder and decoder layer for effective data reconstruction. In the experimental results, we compare the 3D autoencoder-based method with the 2D autoencoder models using the training datasets of 80 healthy subjects from the Human Connectome Project (HCP) and the testing datasets of 25 MS patients from the Longitudinal multiple sclerosis lesion segmentation challenge, and show that the proposed method achieves superior performance in prediction of MS lesion by up to 15%.

• Informatization Policy
• /
• v.24 no.4
• /
• pp.79-92
• /
• 2017

As city functions develop more complex and advanced, interests in smart cities are also increasing. Smart cities refer to the cities effectively solving urban problems such as traffic, safety, welfare, and living issues by utilizing ICT. Recently, many countries are attempting to introduce big data, Internet of Things, and artificial intelligence into smart cities, but they have not yet developed into comprehensive urban services. In this paper, we review the current status of domestic and overseas smart cities and suggest ways to solve issues of data sharing and service compatibility. To this end, we propose a "Deep Learning City Framework" that incorporates the deep learning technology into smart city services, and propose a new smart city strategy that safely shares spatial and temporal data in cities and converges learning data of various cities.

• Korean Journal of Remote Sensing
• /
• v.37 no.1
• /
• pp.111-122
• /
• 2021

Building footprint extraction is an active topic in the domain of remote sensing, since buildings are a fundamental unit of urban areas. Deep convolutional neural networks successfully perform footprint extraction from optical satellite images. However, semantic segmentation produces coarse results in the output, such as blurred and rounded boundaries, which are caused by the use of convolutional layers with large receptive fields and pooling layers. The objective of this study is to generate visually enhanced building objects by directly extracting the vertices of individual buildings by combining instance segmentation and keypoint detection. The target keypoints in building extraction are defined as points of interest based on the local image gradient direction, that is, the vertices of a building polygon. The proposed framework follows a two-stage, top-down approach that is divided into object detection and keypoint estimation. Keypoints between instances are distinguished by merging the rough segmentation masks and the local features of regions of interest. A building polygon is created by grouping the predicted keypoints through a simple geometric method. Our model achieved an F1-score of 0.650 with an mIoU of 62.6 for building footprint extraction using the OpenCitesAI dataset. The results demonstrated that the proposed framework using keypoint estimation exhibited better segmentation performance when compared with Mask R-CNN in terms of both qualitative and quantitative results.

• Journal of the Semiconductor & Display Technology
• /
• v.20 no.4
• /
• pp.6-9
• /
• 2021

The 3D industry is drawing attention for its applications in various markets, including architecture, media, VR/AR, metaverse, imperial broadcast, and etc.. The current feature of the architecture we are introducing is to make 3D models more easily created and modified than conventional ones. Existing methods for generating 3D models mainly obtain values using specialized equipment such as RGB-D cameras and Lidar cameras, through which 3D models are constructed and used. This requires the purchase of equipment and allows the generated 3D model to be verified by the computer. However, our framework allows users to collect data in an easier and cheaper manner using cell phone cameras instead of specialized equipment, and uses 2D data to proceed with 3D modeling on the server and output it to cell phone application screens. This gives users a more accessible environment. In addition, in the 3D modeling process, object classification is attempted through deep learning without user intervention, and mesh and texture suitable for the object can be applied to obtain a lively 3D model. It also allows users to modify mesh and texture through requests, allowing them to obtain sophisticated 3D models.

• Journal of Korea Multimedia Society
• /
• v.25 no.8
• /
• pp.1122-1135
• /
• 2022

With the development of digital technology and the influence of the global pandemic, the metaverse, a three-dimensional virtual world, is receiving attention in society, economy and overall industry, and the manufacturing industry is also accepting it as a major strategic agenda for digital transformation. Therefore, in this study, the concept of the industry metaverse from the perspective of the manufacturing industry was defined, and the types of the industry metaverse were classified into four types by reflecting the characteristics of the manufacturing industry based on the general metaverse scenario presented in previous studies. These are Virtual behavior simulation, Augmented operation of business objects and Virtual experience simulation, Augmented decision of business subjects. In addition, through case analysis of solutions used in the manufacturing industry, it was confirmed that the central technology of the Industry Metaverse is the digital twin, and that it is being implemented by convergence with major digital technologies such as virtual reality, augmented reality, digital human, and AI. This study will be able to provide guidelines for future research on the metaverse from the perspective of the manufacturing industry and establishment of a digital transformation strategy for the industry.

• International Journal of Internet, Broadcasting and Communication
• /
• v.15 no.2
• /
• pp.144-156
• /
• 2023

Smart agriculture is a rapidly growing field that seeks to optimize crop yields and reduce risk through the use of advanced technology. A key challenge in this field is the need to create a comprehensive smart farm system that can effectively monitor and control the growth environment of crops, particularly when cultivating new varieties. This is where fuzzy theory comes in, enabling the collection and analysis of external environmental factors to generate a rule-based system that considers the specific needs of each crop variety. By doing so, the system can easily set the optimal growth environment, reducing trial and error and the user's risk burden. This is in contrast to existing systems where parameters need to be changed for each breed and various factors considered. Additionally, the type of house used affects the environmental control factors for crops, making it necessary to adapt the system accordingly. While developing such a framework requires a significant investment of labour and time, the benefits are numerous and can lead to increased productivity and profitability in the field of smart agriculture. We developed an AI platform for optimal control of facility houses by integrating data from mushroom crops and environmental factors, and analysing the correlation between optimal control conditions and yield. Our experiments demonstrated significant performance improvement compared to the existing system.