• Journal of Information Technology Applications and Management
• /
• 제31권4호
• /
• pp.99-119
• /
• 2024

Focusing on the rapid progress of image generation AI, this study examines the changes in talent required according to changes in the production process of the content industry, and proposes an educational management plan for the subject and comparative department of the university's animation major. First, through environmental analysis, the trend of the animation content industry is analyzed in three stages, and the necessity of producing AI-adapted content talent is derived by re-establishing the talent image of the university's animation major and introducing it into rapid education. Next, we present a case designed by applying teaching methods to improve technology convergence capabilities and project-oriented capabilities by presenting subject and non-curricular cases operated in the animation department of the researcher's university. Through this, we propose the necessity of education to cultivate animation content talent who can play technical and administrative roles by utilizing various AI systems in the future. The goal of this study is to establish a cornerstone study by presenting application cases and having the status of a university as a talent supplier that can lead the content industry beyond the era of AI content production that breaks the boundaries of genres between contents. In conclusion, it is intended to propose the application of education to create value through technology convergence capabilities and project-oriented capabilities to cultivate AI-adapted content talents.

• 생물환경조절학회지
• /
• 제33권2호
• /
• pp.120-128
• /
• 2024

2022년 ChatGPT 출시 이후, 생성형 AI 산업은 엄청난 규모로 성장하였으며, 인지 작업에 혁신을 가져올 것으로 기대되고 있다. 특히 AI 기반 이미지 생성 기술은 현재 디지털 세계의 핵심적인 변화를 주도하고 있다. 본 연구는 대표적인 AI 이미지 생성 도구인 미드저니, 스테이블 디퓨전, 그리고 파이어플라이의 기술적 원리를 분석하고, 이미지 생성 결과를 비교함으로써 그 유용성을 평가하였다. 실험 결과, 이 AI 도구들은 대표 시설원예 작물인 토마토, 딸기, 파프리카, 오이의 과실 이미지를 실제와 유사하게 재현하였다. 특히 파이어플라이는 실제 온실 재배 작물 이미지를 매우 사실적으로 묘사하는 능력을 보여주었다. 그러나 모든 도구들은 작물이 자라는 온실의 환경적 맥락을 완전히 반영하는 데에 있어서 다소 한계를 보였다. 프롬프트 개선 및 레퍼런스 이미지를 활용하여 딸기과실 이미지와 시설 딸기재배 시스템을 보다 정교하게 생성하는 과정도 포함되었으며, 이러한 접근은 AI 이미지 생성 기술의 세밀한 조정이 가능함을 보여준다. 오이 과실 이미지 생성능력을 비교한 결과, AI 생성 도구들은 실제 이미지와 매우 유사한 이미지를 생성해 냄으로써 이미지 생성 점수(CLIP score)에 있어서 통계적 차이를 보이지 않았다. 본 연구는 AI 기반 이미지 생성 이미지 기술이 농업 분야에 활용될 수 있는 방안을 모색하며, 생성형 AI의 농업에 대한 적용을 긍정적으로 전망한다.

• Journal of Information Technology Applications and Management
• /
• 제30권5호
• /
• pp.107-120
• /
• 2023

In this study, we investigate the changes and realities of the content production process focusing on Image generation AI revolutions such as Stable Diffusion, Midjourney, and DELL-E, and examine the current status of related department operations at universities and Find out the status of the current curriculum. Through this, we suggest the need to produce AI-adaptive content talent through re-establishing the capabilities of content-related departments in art universities and quickly introducing curriculum. This is because it can be input into the efficient AI content development system currently being applied in industrial fields, and it is necessary to cultivate talent who can perform managerial and technical roles using various AI systems in the future. In conclusion, we will prepare cornerstone research to establish the university's status as a source of talent that can lead the content industry beyond the AI content production era, and focus on convergence capabilities and experience with the goal of producing convergence talent to cultivate AI adaptive content talent, suggests the direction of curriculum application for value creation.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2024년도 제69차 동계학술대회논문집 32권1호
• /
• pp.481-484
• /
• 2024

본 논문에서는 레고(LEGO®) 조립 설명서를 생성하기 위해 딥러닝을 이용한 조립 및 설명서 생성 시스템을 제안한다. 이 시스템은 사용자가 제공한 단일 이미지를 기반으로 레고 조립 설명서를 자동 생성한다. 해당 시스템은 딥러닝 기반 이미지 분할 기술을 활용하여 물체를 배경으로부터 분리하고 이를 통해 조립 설명서를 생성하는 과정을 포함하며, 조립을 위한 알고리즘을 새로 설계하였다. 이 시스템은 기존 레고 제품의 한계를 극복하고, 사용자에게 주어진 부품으로 다양한 모델을 자유롭게 조립할 수 있게 한다. 또한, 복잡한 레고 조립 과정을 간소화하고, 조립의 장벽을 낮추는 데 도움을 준다.

• Investigative Magnetic Resonance Imaging
• /
• 제26권1호
• /
• pp.1-9
• /
• 2022

Artificial intelligence (AI) powered by deep learning (DL) has shown remarkable progress in image recognition tasks. Over the past decade, AI has proven its feasibility for applications in medical imaging. Various aspects of clinical practice in neuroimaging can be improved with the help of AI. For example, AI can aid in detecting brain metastases, predicting treatment response of brain tumors, generating a parametric map of dynamic contrast-enhanced MRI, and enhancing radiomics research by extracting salient features from input images. In addition, image quality can be improved via AI-based image reconstruction or motion artifact reduction. In this review, we summarize recent clinical applications of DL in various aspects of neuroimaging.

• 문화기술의 융합
• /
• 제9권4호
• /
• pp.667-679
• /
• 2023

이미지 생성형 AI의 등장으로 미술, 디자인 전문가가 아니어도 텍스트 입력을 통해 완성도 높은 그림 작품을 만들 수 있게 되었다. 생성 이미지의 활용 가능성과 예술 산업에 미치는 영향력이 높아짐에 따라 사용자가 AI와 공동 창작하는 과정을 어떻게 인식하는지에 대해 연구 필요성이 제기되고 있다. 이에 본 연구에서는 일반 사용자들을 대상으로 이미지 생성형 AI 창작에 대한 예상 과정과 체감 과정을 알아보고 어떤 과정이 사용자의 창작 주체감에 영향을 미치는지 알아보는 실험 연구를 진행하였다. 연구 결과 사용자들이 기대한 창작 과정과 체감한 창작 과정 간 격차가 있는 것으로 나타났으며 창작 주체감은 낮게 인식하는 경향을 보였다. 이에 AI가 사용자의 창작 의도를 지원하는 조력자의 역할로 작용하여 사용자가 높은 창작 주체감을 경험할 수 있도록 8가지 방법을 제언한다. 본 연구를 통해 사용자 중심적인 창작 경험을 고려하여 향후 이미지 생성형 AI의 발전에 기여할 수 있다.

• International Journal of Internet, Broadcasting and Communication
• /
• 제15권4호
• /
• pp.166-171
• /
• 2023

With significant advancements in the field of artificial intelligence, many novel algorithms and technologies have emerged. Currently, AI painting can generate high-quality images based on textual descriptions. However, it is often challenging to control details when generating images, even with complex textual inputs. Therefore, there is a need to implement additional control mechanisms beyond textual descriptions. Based on ControlNet, this passage describes a combined utilization of various local controls (such as edge maps and depth maps) and global control within a single model. It provides a comprehensive exposition of the fundamental concepts of ControlNet, elucidating its theoretical foundation and relevant technological features. Furthermore, combining methods and applications, understanding the technical characteristics involves analyzing distinct advantages and image differences. This further explores insights into the development of image generation patterns.

• International Journal of Internet, Broadcasting and Communication
• /
• 제16권1호
• /
• pp.86-91
• /
• 2024

In the field of artistic creation, the iterative development of AI-generated video software has pushed the boundaries of multimedia content creation and provided powerful creative tools for non-professionals. This paper extensively examines two leading AI-generated video software, PIKA Labs and RUNWAY, discussing their functions, performance differences, and application scopes in the video generation domain. Through detailed operational examples, a comparative analysis of their functionalities, as well as the advantages and limitations of each in generating video content, is presented. By comparison, it can be found that PIKA Labs and RUNWAY have excellent performance in stability and creativity. Therefore, the purpose of this study is to comprehensively elucidate the operating mechanisms of these two AI software, in order to intuitively demonstrate the advantages of each software. Simultaneously, this study provides valuable references for professionals and creators in the video production field, assisting them in selecting the most suitable tools for different scenarios, thereby advancing the application and development of AI-generated video software in multimedia content creation.

• 한국BIM학회 논문집
• /
• 제14권2호
• /
• pp.13-24
• /
• 2024

This research suggests a novel visualization approach utilizing Generative AI to render photorealistic architectural alternatives images in the early design phase. Photorealistic rendering intuitively describes alternatives and facilitates clear communication between stakeholders. Nevertheless, the conventional rendering process, utilizing 3D modelling and rendering engines, demands sophisticate model and processing time. In this context, the paper suggests a rendering approach employing the text-to-image method aimed at generating a broader range of intuitive and relevant reference images. Additionally, it employs an Text-to-Image method focused on producing a diverse array of alternatives reflecting architects' styles when visualizing the exteriors of residential buildings from the mass model images. To achieve this, fine-tuning for architects' styles was conducted using the Low-Rank Adaptation (LoRA) method. This approach, supported by fine-tuned models, allows not only single style-applied alternatives, but also the fusion of two or more styles to generate new alternatives. Using the proposed approach, we generated more than 15,000 meaningful images, with each image taking only about 5 seconds to produce. This demonstrates that the Generative AI-based visualization approach significantly reduces the labour and time required in conventional visualization processes, holding significant potential for transforming abstract ideas into tangible images, even in the early stages of design.

• 전자통신동향분석
• /
• 제38권5호
• /
• pp.1-11
• /
• 2023

The emergence of generative hyperscale artificial intelligence (AI) has enabled new services, such as image-generating AI and conversational AI based on large language models. Such services likely lead to the influx of numerous users, who cannot be handled using conventional AI models. Furthermore, the exponential increase in training data, computations, and high user demand of AI models has led to intensive hardware resource consumption, highlighting the need to develop domain-specific semiconductors for hyperscale AI. In this technical report, we describe development trends in technologies for hyperscale AI processors pursued by domestic and foreign semiconductor companies, such as NVIDIA, Graphcore, Tesla, Google, Meta, SAPEON, FuriosaAI, and Rebellions.