• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.23-26
• /
• 2022

딥 러닝 모델은 블랙 박스 (Black Box) 모델로 예측에 대한 근거를 제시하지 못해 신뢰성이 떨어지는 단점이 존재한다. 이를 해결하기 위해 딥 러닝 모델에 설명력을 부여하는 설명 가능한 인공지능 (XAI) 분야 연구가 활발하게 이루어지고 있다. 본 논문에서는 모델 예측을 프로토타입을 통해 설명하는 딥 러닝 모델을 제시한다. 즉, "주어진 이미지는 티셔츠인데, 그 이유는 티셔츠를 대표하는 모양의 프로토타입과 닮았기 때문이다."의 형태로 딥 러닝 모델을 설명한다. 해당 모델은 Encoder, Prototype Layer, Classifier로 구성되어 있다. Encoder는 Feature를 추출하는 데 활용하고 Classifier를 통해 분류 작업을 수행한다. 모델이 제시하는 분류 결과를 설명하기 위해 Prototype Layer에서 가장 유사한 프로토타입을 찾아 설명을 제시한다. 실험 결과 프로토타입 생성 기반 설명 모델은 기존 이미지 분류 모델과 유사한 예측 정확도를 보였고, 예측에 대한 설명력까지 확보하였다.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.29 no.5
• /
• pp.1019-1026
• /
• 2019

Due to limitations of existing methods for detecting newly introduced malware, the importance of the development of artificial intelligence vaccines arises. Existing artificial intelligence vaccines have a disadvantage that the accuracy of the detection rate is low because those vaccines do not scan all parts of the file. In this paper, we suggest an enhanced method for detecting malware which is composed of unique OP Code features in the malware files. Specifically, we tested the method with text datasets trained on Random Forest algorithm and with image datasets trained on the Inception V3 model. As a result, the highest accuracy of the detection rate was about 80%.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.550-552
• /
• 2022

최근 사물인터넷(IoT) 기기가 활성화됨에 따라 웨어러블 장치 환경에서 장기간 모니터링 및 수집이 가능해짐에 따라 생체 신호 처리 및 ECG 분석 연구가 활성화되고 있다. 그러나, ECG 데이터는 부정맥 비트의 불규칙적인 발생으로 인한 클래스 불균형 문제와 근육의 떨림 및 신호의 미약등과 같은 잡음으로 인해 낮은 신호 품질이 발생할 수 있으며 훈련용 공개데이터 세트가 작다는 특징을 갖는다. 이 논문에서는 ECG 1D 신호를 2D 스펙트로그램 이미지로 변환하여 잡음의 영향을 최소화하고 전이학습과 메타학습의 장점을 결합하여 클래스 불균형 문제와 소수의 데이터에서도 빠른 학습이 가능하다는 특징을 갖는다. 따라서, 이 논문에서는 ECG 스펙트럼 이미지를 사용하여 2D-CNN 메타-전이 학습 기반 부정맥 분류 기법을 제안한다.

• Smart Media Journal
• /
• v.12 no.4
• /
• pp.9-18
• /
• 2023

Steganography is a technology capable of protecting data by hiding the existence of data. Recently, with the development of deep learning technology, deep learning-based steganography are being developed. Deep learning can learn by analyzing high-dimensional features of data, so it can improve the performance and quality of steganography. In this paper, we investigated the research trend of image steganography based on deep learning.

• Journal of Platform Technology
• /
• v.10 no.1
• /
• pp.20-28
• /
• 2022

Artificial intelligence-based health data verification has become an essential element not only to help clinical research, but also to develop new treatments. Since the US Food and Drug Administration (FDA) approved the marketing of medical devices that detect mild abnormal diabetic retinopathy in adult diabetic patients using artificial intelligence in the field of medical diagnosis, tests using artificial intelligence have been increasing. In this study, an artificial intelligence model based on image classification was created using a Teachable Machine supported by Google, and a predictive model was completed through learning. This not only facilitates the early detection of cataracts among eye diseases occurring among patients with chronic diseases, but also serves as basic research for developing a digital personal health healthcare app for eye disease prevention as a healthcare program for eye health.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.520-522
• /
• 2022

의료 인공지능은 특정 진단에서 높은 정확도를 보이지만 모델의 신뢰성 문제로 인해 활발하게 쓰이지 못하고 있다. 이에 따라 인공지능 모델의 진단에 대한 원인 설명의 필요성이 대두되었고 설명가능한 의료 인공지능에 관한 연구가 활발히 진행되고 있다. 하지만 MRI 등 의료 영상 인공지능 분야에서 주로 진행되고 있으며, 이미지 형태가 아닌 전자의무기록 데이터 (Electronic Health Record, EHR) 를 기반으로 한 모델의 설명가능성 연구는 EHR 데이터 자체의 복잡성 때문에 활발하게 진행 되지 않고 있다. 본 논문에서는 전자의무기록 데이터인 MIMIC-III (Medical Information Mart for Intensive Care) 를 전처리 및 그래프로 표현하고, GCT (Graph Convolutional Transformer) 모델을 학습시켰다. 학습 후, 어텐션 흐름 그래프를 시각화해서 모델의 예측에 대한 직관적인 설명을 제공한다.

• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2022.04a
• /
• pp.225-226
• /
• 2022

The construction industry, which has a labour-intensive and conservative nature, is exclusive to adopt new technologies. However, the construction industry is viably introducing the 4th Industrial Revolution technologies represented by artificial intelligence, Internet of Things, robotics and unmanned transportation to promote change into a smart industry. An image-based artificial intelligence technology is a field of computer vision technology that refers to machines mimicking human visual recognition of objects from pictures or videos. The purpose of this article is to explore image-based artificial intelligence technologies which would be able to apply to the construction sites. In this study, we show two examples which is one for a construction waste classification model and another for cast in-situ anchor bolts defection detection model. Image-based intelligence technologies would be used for various measurement, classification, and detection works that occur in the construction projects.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.8
• /
• pp.347-354
• /
• 2022

Since the era of the 4th industrial revolution has been ripe, the use of artificial intelligence(AI) based on massive data is beginning to be actively applied in various fields. However, as the process of analyzing forest species is carried out manually, many errors are occurring. Therefore, in this paper, about 60,000 pieces of AI learning data were automatically analyzed for pine, larch, conifer, and broadleaf trees of aerial photographs and pseudo images in the metropolitan area, and an AI model was developed to distinguish tree species. Through this, it is expected to increase in work efficiency by using the tree species division image as basic data when producing forest change detection and forest field topics.

• Journal of the Korean Society of Radiology
• /
• v.17 no.6
• /
• pp.873-879
• /
• 2023

This study explored the use of artificial intelligence(AI) to detect foreign bodies in chest X-ray images. Medical imaging, especially chest X-rays, plays a crucial role in diagnosing diseases such as pneumonia and lung cancer. With the increase in imaging tests, AI has become an important tool for efficient and fast diagnosis. However, images can contain foreign objects, including everyday jewelry like buttons and bra wires, which can interfere with accurate readings. In this study, we developed an AI algorithm that accurately identifies these foreign objects and processed the National Institutes of Health chest X-ray dataset based on the YOLOv8 model. The results showed high detection performance with accuracy, precision, recall, and F1-score all close to 0.91. Despite the excellent performance of AI, the study solved the problem that foreign objects in the image can distort the reading results, emphasizing the innovative role of AI in radiology and its reliability based on accuracy, which is essential for clinical implementation.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.6
• /
• pp.1087-1098
• /
• 2023

The rapid advancement of artificial intelligence (AI) technology has led to its proactive utilization across various fields. However, this widespread adoption of AI-based systems has raised concerns about the increasing threat of attacks on these systems. In particular, deep neural networks, commonly used in deep learning, have been found vulnerable to adversarial attacks that intentionally manipulate input data to induce model errors. In this study, we propose a method to protect image classification models from visually imperceptible One-Pixel attacks, where only a single pixel is altered in an image. The proposed defense technique utilizes an autoencoder model to remove potential threat elements from input images before forwarding them to the classification model. Experimental results, using the CIFAR-10 dataset, demonstrate that the autoencoder-based defense approach significantly improves the robustness of pretrained image classification models against One-Pixel attacks, with an average defense rate enhancement of 81.2%, all without the need for modifications to the existing models.