• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권4호
• /
• pp.1704-1720
• /
• 2020

Recognizing food from photographs presents many applications for machine learning, computer vision and dietetics, etc. Recent progress of deep learning techniques accelerates the recognition of food in a great scale. We build a hierarchical structure composed of deep CNN to recognize and classify food from photographs. We build a dataset for Korean food of 18 classes, which are further categorized in 4 major classes. Our hierarchical recognizer classifies foods into four major classes in the first step. Each food in the major classes is further classified into the exact class in the second step. We employ DenseNet structure for the baseline of our recognizer. The hierarchical structure provides higher accuracy and F1 score than those from the single-structured recognizer.

• 전기전자학회논문지
• /
• 제21권2호
• /
• pp.136-141
• /
• 2017

본 논문에서는 딥러닝을 PC에 적용하기 위한 메모리 최적화에 관한 알고리즘을 제안한다. 제안된 알고리즘은 일반 PC에서 기존의 딥러닝 구조에서 요구되는 연산처리 과정과 데이터 량을 감소시켜 메모리 및 연산처리 시간을 최소화한다. 본 논문에서 제안하는 알고리즘은 분별력이 있는 랜덤 필터를 이용한 컨볼루션 층 구성 과정, PCA를 이용한 데이터 축소 과정, SVM을 사용한 CNN 구조 생성 등의 3과정으로 이루어진다. 분별력이 있는 랜덤 필터를 이용한 컨볼루션 층 구성 과정에서는 학습과정이 필요치 않아서 전체적인 딥러닝의 학습시간을 단축시킨다. PCA를 이용한 데이터 축소 과정에서는 메모리량과 연산처리량을 감소시킨다. SVM을 사용한 CNN 구조 생성에서는 필요로 하는 메모리량과 연산 처리량의 감소 효과를 극대화 시킨다. 제안된 알고리즘의 성능을 평가하기 위하여 예일 대학교의 Extended Yale B 얼굴 데이터베이스를 사용하여 실험한 결과, 본 논문에서 제안하는 알고리즘이 기존의 CNN 알고리즘과 비교하여 비슷한 성능의 인식률을 보이면서 연산 소요시간과 메모리 점유율에 있어 우수함이 확인되었다. 본 논문에서 제안한 알고리즘을 바탕으로 하여 일반 PC에서도 많은 데이터와 연산처리를 가진 딥러닝 알고리즘을 구현할 수 있으리라 기대된다.

• 대한토목학회논문집
• /
• 제40권2호
• /
• pp.239-245
• /
• 2020

최근 BIM (Building Information Modeling)이 건설 산업계에서 폭넓게 활용되고 있다. 하지만 과거에 시공이 된 구조물에 경우 대부분 BIM이 구축되어 있지 않다. BIM이 구축되지 않은 구조물의 경우, 카메라로부터 얻은 2D 이미지에 SfM (Structure from Motion) 기법을 활용하면 3D 모델의 점군 데이터(Point cloud)를 생성하고 BIM을 구축할 수 있다. 하지만 이렇게 생성된 점군 데이터는 의미론적 정보가 포함되어 있지 않기 때문에, 수작업으로 구조물의 어떤 요소인지 분류해 주어야 한다. 따라서 본 연구에서는 구조물 구성요소를 분류하는 과정을 자동화하기 위하여 딥러닝을 적용하였다. 딥러닝 네트워크 구축에는 CNN (Convolutional Neural Network) 구조의 Inception-ResNet-v2를 사용하였고, 전이학습을 통하여 교량 구조물의 구성요소를 학습하였다. 개발된 시스템을 검증하기 위하여 수집한 데이터를 이용하여 구성요소를 분류한 결과, 교량의 구성요소를 96.13 %의 정확도로 분류할 수 있었다.

• 한국컴퓨터정보학회논문지
• /
• 제24권1호
• /
• pp.9-23
• /
• 2019

In this paper, we propose a layer structure of a pest image classifier model using CNN (Convolutional Neural Network) and background removal image processing algorithm for improving classification accuracy in order to build a smart monitoring system for pine wilt pest control. In this study, we have constructed and trained a CNN classifier model by collecting image data of pine wilt pest mediators, and experimented to verify the classification accuracy of the model and the effect of the proposed classification algorithm. Experimental results showed that the proposed method successfully detected and preprocessed the region of the object accurately for all the test images, resulting in showing classification accuracy of about 98.91%. This study shows that the layer structure of the proposed CNN classifier model classified the targeted pest image effectively in various environments. In the field test using the Smart Trap for capturing the pine wilt pest mediators, the proposed classification algorithm is effective in the real environment, showing a classification accuracy of 88.25%, which is improved by about 8.12% according to whether the image cropping preprocessing is performed. Ultimately, we will proceed with procedures to apply the techniques and verify the functionality to field tests on various sites.

• 한국멀티미디어학회논문지
• /
• 제21권9호
• /
• pp.1062-1067
• /
• 2018

Conventional active sonar technology has relied heavily on the hearing of sonar operator, but recently, many techniques for automatic detection and classification have been studied. In this paper, we extract the image data from the spectrogram of the active sonar signal and classify the extracted data using CNN(convolutional neural networks), which has recently presented excellent performance improvement in the field of pattern recognition. First, we divided entire data set into eight classes depending on the ratio containing the target. Then, experiments were conducted to classify the eight classes data using proposed CNN structure, and the results were analyzed.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 추계학술대회 논문집 학회본부 B
• /
• pp.651-653
• /
• 1998

This thesis propose the prefilter type control strategies using modified chaotic neural networks #or the trajectory control of robotic manipulator. Since the structure of chaotic neural networks and neurons, chaotic neural networks can show the robust characteristics for controlling highly nonlinear dynamics like robotic manipulators. For its application, the trajectory controller of the three-axis PUMA robot is designed by CNN. The CNN controller acts as the compensator of the PD controller. Simulation results show that learning error decrease drastically via on- line learning and the performance is excellent. The CNN controller have much better controllability and shorter calculation time compared to the RNN controller. Another advantage of the proposed controller could be attached to conventional robot controller without hardware changes.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권5호
• /
• pp.2558-2573
• /
• 2019

We present a deep CNN-based approach for classifying artistic media from artwork images. We aim to classify most frequently used artistic media including oilpaint brush, watercolor brush, pencil and pastel, etc. For this purpose, we extend VGGNet, one of the most widely used CNN structure, by substituting its last layer with a fully convolutional layer, which reveals class activation map (CAM), the region of classification. We build two artwork image datasets: YMSet that collects more than 4K artwork images for four most frequently used artistic media from various internet websites and WikiSet that collects almost 9K artwork images for ten most frequently used media from WikiArt. We execute a human baseline experiment to compare the classification performance. Through our experiments, we conclude that our classifier is superior in classifying artistic media to human.

• 한국멀티미디어학회논문지
• /
• 제24권5호
• /
• pp.701-708
• /
• 2021

As one of the most widely used technology in the world right now, Face recognition has already received widespread attention by all the researcher and institutes. It has been used in many fields such as safety protection, surveillance system, crime control and even in our ordinary life such as home security and so on. This technology with today's technology has advantages such as high connectivity and real time transformation. But we still need to improve its recognition rate, reaction time and also reduce impact of different environmental status to the whole system. So in this paper we proposed a face recognition system model with improved CNN which combining the characteristics of flat network and residual network, integrated learning, simplify network structure and enhance portability and also improve the recognition accuracy. We also used AR and ORL database to do the experiment and result shows higher recognition rate, efficiency and robustness for different image conditions.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2022년도 추계학술대회
• /
• pp.112-115
• /
• 2022

The process of deep learning usually needs to deal with massive data which has greatly limited the development of deep learning technologies today. Convolutional Neural Network (CNN) structure is often used to solve image classification problems. However, a large number of images may be required in order to train an image in CNN, which is a heavy burden for existing computer systems to handle. If the image data can be compressed under the premise that the computer hardware system remains unchanged, it is possible to train more datasets in deep learning. However, image compression usually adopts the form of lossy compression, which will lose part of the image information. If the lost information is key information, it may affect learning performance. In this paper, we will analyze the effect of image compression on deep learning performance on CNN-based cat and dog classification. Through the experiment results, we conclude that the compression of images does not have a significant impact on the accuracy of deep learning.

• International Journal of Advanced Culture Technology
• /
• 제11권1호
• /
• pp.325-331
• /
• 2023

In this paper, we propose a mixed neural network structure of CNN and LSTM that can be used to detect or predict odor occurrence, which is most required in manufacturing industry or real life, using odor complex sensors. In addition, the proposed learning model uses a complex odor sensor to receive four types of data such as hydrogen sulfide, ammonia, benzene, and toluene in real time, and applies this data to an inference model to detect and predict odor conditions. The proposed model evaluated the prediction accuracy of the learning model through performance indicators according to accuracy, and the evaluation result showed an average performance of 94% or more.