• 스마트미디어저널
• /
• 제6권3호
• /
• pp.49-56
• /
• 2017

Images and Videos that include the human face contain a lot of information. Therefore, accurately extracting human face is a very important issue in the field of computer vision. However, in real life, human faces have various shapes and textures. To adapt to these variations, A model-based approach is one of the best ways in which unknown data can be represented by the model in which it is built. However, the model-based approach has its weaknesses when the motion between two frames is big, it can be either a sudden change of pose or moving with fast speed. In this paper, we propose an enhanced human face-tracking model. This approach included human face detection and motion estimation using Cascaded Convolutional Neural Networks, and continuous human face tracking and modeling correction steps using the Active Appearance Model. A proposed system detects human face in the first input frame and initializes the models. On later frames, Cascaded CNN face detection is used to estimate the target motion such as location or pose before applying the old model and fit new target.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2023년도 제67차 동계학술대회논문집 31권1호
• /
• pp.369-372
• /
• 2023

본 논문에서는 삼각형 구조로 구성된 3차원 메쉬(Mesh)에서 합성곱 신경망(Convolution Neural Network, CNN)을 응용하여 정확도가 높은 새로운 학습 표현 기법을 제시한다. 우리는 메쉬를 구성하고 있는 폴리곤의 edge와 face의 로컬 특징을 기반으로 학습을 진행한다. 일반적으로 딥러닝은 인공신경망을 수많은 계층 형태로 연결한 기법을 말하며, 주요 처리 대상은 1, 2차원 데이터 형태인 오디오 파일과 이미지였다. 인공지능에 대한 연구가 지속되면서 3차원 딥러닝이 도입되었지만, 기존의 학습과는 달리 3차원 딥러닝은 데이터의 확보가 쉽지 않다. 혼합현실과 메타버스 시장의 확대로 인해 3차원 모델링 시장이 증가하고, 기술의 발전으로 데이터를 획득할 수 있는 방법이 생겼지만, 3차원 데이터를 직접적으로 학습에 이용하는 방식으로 적용하는 것은 쉽지 않다. 그렇게 때문에 본 논문에서는 산업 현장에서 이용되는 데이터인 메쉬 구조를 폴리곤의 최소 단위인 삼각형 형태로 구성하여 학습 데이터를 구성해 기존의 방법보다 정확도가 높은 학습 기법을 제안한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권11호
• /
• pp.2903-2923
• /
• 2023

Knowledge recommendation is a type of recommendation system that recommends knowledge content to users in order to satisfy their needs. Although using graph neural networks to extract data features is an effective method for solving the recommendation problem, there is information loss when modeling real-world problems because an edge in a graph structure can only be associated with two nodes. Because one super-edge in the hypergraph structure can be connected with several nodes and the effectiveness of knowledge graph for knowledge expression, a dual-channel hypergraph convolutional neural network model (DCHC) based on hypergraph structure and knowledge graph is proposed. The model divides user data and knowledge data into user subhypergraph and knowledge subhypergraph, respectively, and extracts user data features by dual-channel hypergraph convolution and knowledge data features by combining with knowledge graph technology, and finally generates recommendation results based on the obtained user embedding and knowledge embedding. The performance of DCHC model is higher than the comparative model under AUC and F1 evaluation indicators, comparative experiments with the baseline also demonstrate the validity of DCHC model.

• Proceedings of the National Institute of Ecology of the Republic of Korea
• /
• 제4권4호
• /
• pp.159-176
• /
• 2023

The conservation of the raccoon dog (Nyctereutes procyonoides) in South Korea requires the protection and preservation of natural habitats while additionally ensuring coexistence with human activities. Applying habitat map modeling techniques provides information regarding the distributional patterns of raccoon dogs and assists in the development of future conservation strategies. The purpose of this study is to generate potential habitat distribution maps for the raccoon dog in South Korea using geospatial technology-based models. These models include the frequency ratio (FR) as a bivariate statistical approach, the group method of data handling (GMDH) as a machine learning algorithm, and convolutional neural network (CNN) and long short-term memory (LSTM) as deep learning algorithms. Moreover, the imperialist competitive algorithm (ICA) is used to fine-tune the hyperparameters of the machine learning and deep learning models. Moreover, there are 14 habitat characteristics used for developing the models: elevation, slope, valley depth, topographic wetness index, terrain roughness index, slope height, surface area, slope length and steepness factor (LS factor), normalized difference vegetation index, normalized difference water index, distance to drainage, distance to roads, drainage density, and morphometric features. The accuracy of prediction is evaluated using the area under the receiver operating characteristic curve. The results indicate comparable performances of all models. However, the CNN demonstrates superior capacity for prediction, achieving accuracies of 76.3% and 75.7% for the training and validation processes, respectively. The maps of potential habitat distribution are generated for five different levels of potentiality: very low, low, moderate, high, and very high.

• 응용통계연구
• /
• 제37권2호
• /
• pp.139-155
• /
• 2024

본 연구는 좌측 해마 경화를 보인 내측두엽 뇌전증(left mTLE, mesial temporal lobe epilepsy with left hippocampal sclerosis) 환자군과 우측 해마 경화를 보인 내측두엽 뇌전증(right mTLE, mesial temporal lobe epilepsy with right hippocampal sclerosis) 환자군 그리고 건강한 대조군(healthy controls; HC)으로부터 측정한 뇌자도(magnetoencephalography; MEG) 데이터로 각 그룹을 분류하는 다중 분류 작업에 다양한 인공신경망을 적용하고 그 결과를 비교해 보고자 하였다. 합성곱 신경망, 순환 신경망 그리고 그래프 신경망으로 모델링한 결과, k-fold 정확도 평균은 합성곱 신경망 기반 모델, 그래프 신경망 기반 모델, 순환 신경망 기반 모델 순으로 우수하였다. 또한, 수행 시간은 순환 신경망 기반 모델, 그래프 신경망 기반 모델, 합성곱 신경망 기반 모델 순으로 우수하였다. 정확도 성능과 시간 면에서 모두 좋은 수치를 보이며, 네트워크 데이터의 확장성이 뛰어난 그래프 신경망이 앞으로 뇌 연구에 활용되기 적합한 모델임을 강조하고자 한다.

• 전자공학회논문지
• /
• 제54권3호
• /
• pp.70-77
• /
• 2017

최근 IoT 및 딥러닝 관련 기술요소들이 영상보안감시시스템에서도 다양하게 응용되고 있다. 그 중 CCTV를 통해 촬영된 동영상에서 자동으로 특정 객체를 검출, 추적, 분류 하는 감시 기능이 점점 지능화되고 있다. 본 논문에서는 보급형 CPU만 사용하는 PC 환경에서도 실시간 처리가 가능한 알고리즘을 목표로 하였다. GMM(Gaussian Mixture Model)을 이용한 배경 모델링과 헝가리안 알고리즘, 그리고 칼만 필터를 조합한 추적 알고리즘은 전통적이며 복잡도가 비교적 적지만 검출 오류가 높다. 이를 보강하기 위해 대용량 데이터 학습에 적합한 딥러닝을 기술을 적용하였다. 특히 움직임이 있는 사람의 특징을 강조하기 위해 추적된 객체에 대해 SRGB-3 Layer CNN을 사용하였다. 성능 평가를 위해 기존의 HOG와 SVM을 이용한 시스템과 비교했을 때 Move-in은 7.6%, Move-out은 9.0%의 오류율 감소가 있었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권5호
• /
• pp.1396-1412
• /
• 2023

Conversation modeling is an important and challenging task in the field of natural language processing because it is a key component promoting the development of automated humanmachine conversation. Most recent research concerning conversation modeling focuses only on the current utterance (considered as the current question) to generate a response, and thus fails to capture the conversation's logic from its beginning. Some studies concatenate the current question with previous conversation sentences and use it as input for response generation. Another approach is to use an encoder to store all previous utterances. Each time a new question is encountered, the encoder is updated and used to generate the response. Our approach in this paper differs from previous studies in that we explicitly separate the encoding of the question from the encoding of its context. This results in different encoding models for the question and the context, capturing the specificity of each. In this way, we have access to the entire context when generating the response. To this end, we propose a deep neural network-based model, called the Context Model, to encode previous utterances' information and combine it with the current question. This approach satisfies the need for context information while keeping the different roles of the current question and its context separate while generating a response. We investigate two approaches for representing the context: Long short-term memory and Convolutional neural network. Experiments show that our Context Model outperforms a baseline model on both ConvAI2 Dataset and a collected dataset of conversational English.

• 한국정보전자통신기술학회논문지
• /
• 제13권4호
• /
• pp.283-292
• /
• 2020

4차 산업혁명의 대표적인 이미지 중 하나인 인공지능은 2016년 알파고 이후에 인공지능 인식이 매우 높아져 있다. 본 논문은 학국교육학술정보원에서 제공하는 국내 논문 중 '인공지능', '기계학습', '딥 러닝'으로 검색된 국내 발표 논문에 대해서 분석하였다. 검색된 논문은 약 1만여건이며 논문 동향을 파악하기 위해 빈도분석과 토픽 모델링, 의미 연결망을 이용하였다. 추출된 논문을 분석한 결과, 2015년에 비해 2016년에는 인공지능 분야는 600%, 기계학습은 176%, 딥 러닝 분야는 316% 증가하여 알파고 이후에 인공지능 분야의 연구가 활발히 진행됨을 확인할 수 있었다. 또한, 2018년 부터는 기계학습보다 딥 러닝 분야가 더 많이 연구 발표되고 있다. 기계학습에서는 서포트 벡터 머신 모델이, 딥 러닝에서는 텐서플로우를 이용한 컨볼루션 신경망이 많이 활용되고 있음을 알 수 있었다. 본 논문은 '인공지능', '기계학습', '딥 러닝' 분야의 향후 연구 방향을 설정하는 도움을 제공할 수 있다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.681-684
• /
• 2021

In this paper, we study image clustering without labeling using machine learning techniques. We proposed an unsupervised machine learning technique to design an image clustering model that automatically categorizes images into groups. Our experiment focused on inception convolutional neural networks (inception V3) with k-mean methods to cluster images. For this, we collect the public datasets containing Food-K5, Flowers, Handwritten Digit, Cats-dogs, and our dataset Rice Germination, and the owner dataset Palm print. Our experiment can expand into three-part; First, format all the images to un-label and move to whole datasets. Second, load dataset into the inception V3 extraction image features and transferred to the k-mean cluster group hold on six classes. Lastly, evaluate modeling accuracy using the confusion matrix base on precision, recall, F1 to analyze. In this our methods, we can get the results as 1) Handwritten Digit (precision = 1.000, recall = 1.000, F1 = 1.00), 2) Food-K5 (precision = 0.975, recall = 0.945, F1 = 0.96), 3) Palm print (precision = 1.000, recall = 0.999, F1 = 1.00), 4) Cats-dogs (precision = 0.997, recall = 0.475, F1 = 0.64), 5) Flowers (precision = 0.610, recall = 0.982, F1 = 0.75), and our dataset 6) Rice Germination (precision = 0.997, recall = 0.943, F1 = 0.97). Our experiment showed that modeling could get an accuracy rate of 0.8908; the outcomes state that the proposed model is strongest enough to differentiate the different images and classify them into clusters.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제18권3호
• /
• pp.755-778
• /
• 2024

In recent years, the number of devices being connected to the internet has grown enormously, as has the intrusive behavior in the network. Thus, it is important for intrusion detection systems to report all intrusive behavior. Using deep learning and machine learning algorithms, intrusion detection systems are able to perform well in identifying attacks. However, the concern with these deep learning algorithms is their inability to identify a suitable network based on traffic volume, which requires manual changing of hyperparameters, which consumes a lot of time and effort. So, to address this, this paper offers a solution using the extended compact genetic algorithm for the automatic tuning of the hyperparameters. The novelty in this work comes in the form of modeling the problem of identifying attacks as a multi-objective optimization problem and the usage of linkage learning for solving the optimization problem. The solution is obtained using the feature map-based Convolutional Neural Network that gets encoded into genes, and using the extended compact genetic algorithm the model is optimized for the detection accuracy and latency. The CIC-IDS-2017 and 2018 datasets are used to verify the hypothesis, and the most recent analysis yielded a substantial F1 score of 99.23%. Response time, CPU, and memory consumption evaluations are done to demonstrate the suitability of this model in a fog environment.