딥러닝의 모형과 응용사례

안성만(Ahn, SungMahn) 한국지능정보시스템학회 2016 지능정보연구 Vol.22 No.2

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for “backward propagation of errors” and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer’s) neurons. Shared weights mean that we’re going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren’t just propagated backward through layers, they’re propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when traini

그래프 합성곱-신경망 구조 탐색 : 그래프 합성곱 신경망을 이용한 신경망 구조 탐색

최수연,박종열 국제문화기술진흥원 2023 The Journal of the Convergence on Culture Technolo Vol.9 No.1

This paper proposes the design of a neural network structure search model using graph convolutional neural networks. Deep learning has a problem of not being able to verify whether the designed model has a structure with optimized performance due to the nature of learning as a black box. The neural network structure search model is composed of a recurrent neural network that creates a model and a convolutional neural network that is the generated network. Conventional neural network structure search models use recurrent neural networks, but in this paper, we propose GC-NAS, which uses graph convolutional neural networks instead of recurrent neural networks to create convolutional neural network models. The proposed GC-NAS uses the Layer Extraction Block to explore depth, and the Hyper Parameter Prediction Block to explore spatial and temporal information (hyper parameters) based on depth information in parallel. Therefore, since the depth information is reflected, the search area is wider, and the purpose of the search area of the model is clear by conducting a parallel search with depth information, so it is judged to be superior in theoretical structure compared to GC-NAS. GC-NAS is expected to solve the problem of the high-dimensional time axis and the range of spatial search of recurrent neural networks in the existing neural network structure search model through the graph convolutional neural network block and graph generation algorithm. In addition, we hope that the GC-NAS proposed in this paper will serve as an opportunity for active research on the application of graph convolutional neural networks to neural network structure search. 본 논문은 그래프 합성곱 신경망을 이용한 신경망 구조 탐색 모델 설계를 제안한다. 딥 러닝은 블랙박스로 학습이 진행되는 특성으로 인해 설계한 모델이 최적화된 성능을 가지는 구조인지 검증하지 못하는 문제점이 존재한다. 신경망 구조 탐색 모델은 모델을 생성하는 순환 신경망과 생성된 네트워크인 합성곱 신경망으로 구성되어있다. 통상의 신경망 구조 탐색 모델은 순환신경망 계열을 사용하지만 우리는 본 논문에서 순환신경망 대신 그래프 합성곱 신경망을 사용하여 합성곱 신경망 모델을 생성하는 GC-NAS를 제안한다. 제안하는 GC-NAS는 Layer Extraction Block을 이용하여 Depth를 탐색하며 Hyper Parameter Prediction Block을 이용하여 Depth 정보를 기반으로 한 spatial, temporal 정보(hyper parameter)를 병렬적으로 탐색합니다. 따라서 Depth 정보를 반영하기 때문에 탐색 영역이 더 넓으며 Depth 정보와 병렬적 탐색을 진행함으로 모델의 탐색 영역의 목적성이 분명하기 때문에 GC-NAS대비 이론적 구조에 있어서 우위에 있다고 판단된다. GC-NAS는 그래프 합성곱 신경망 블록 및 그래프 생성 알고리즘을 통하여 기존 신경망 구조 탐색 모델에서 순환 신경망이 가지는 고차원 시간 축의 문제와 공간적 탐색의 범위 문제를 해결할 것으로 기대한다. 또한 우리는 본 논문이 제안하는 GC-NAS를 통하여 신경망 구조 탐색에 그래프 합성곱 신경망을 적용하는 연구가 활발히 이루어질 수 있는 계기가 될 수 있기를 기대한다.

Deep Neural Network와 Convolutional Neural Network 모델을 이용한 산사태 취약성 매핑

공성현,백원경,정형섭,Gong, Sung-Hyun,Baek, Won-Kyung,Jung, Hyung-Sup 대한원격탐사학회 2022 大韓遠隔探査學會誌 Vol.38 No.6

Landslides are one of the most prevalent natural disasters, threating both humans and property. Also landslides can cause damage at the national level, so effective prediction and prevention are essential. Research to produce a landslide susceptibility map with high accuracy is steadily being conducted, and various models have been applied to landslide susceptibility analysis. Pixel-based machine learning models such as frequency ratio models, logistic regression models, ensembles models, and Artificial Neural Networks have been mainly applied. Recent studies have shown that the kernel-based convolutional neural network (CNN) technique is effective and that the spatial characteristics of input data have a significant effect on the accuracy of landslide susceptibility mapping. For this reason, the purpose of this study is to analyze landslide vulnerability using a pixel-based deep neural network model and a patch-based convolutional neural network model. The research area was set up in Gangwon-do, including Inje, Gangneung, and Pyeongchang, where landslides occurred frequently and damaged. Landslide-related factors include slope, curvature, stream power index (SPI), topographic wetness index (TWI), topographic position index (TPI), timber diameter, timber age, lithology, land use, soil depth, soil parent material, lineament density, fault density, normalized difference vegetation index (NDVI) and normalized difference water index (NDWI) were used. Landslide-related factors were built into a spatial database through data preprocessing, and landslide susceptibility map was predicted using deep neural network (DNN) and CNN models. The model and landslide susceptibility map were verified through average precision (AP) and root mean square errors (RMSE), and as a result of the verification, the patch-based CNN model showed 3.4% improved performance compared to the pixel-based DNN model. The results of this study can be used to predict landslides and are expected to serve as a scientific basis for establishing land use policies and landslide management policies.

GPGPU 기반 Convolutional Neural Network의 효율적인 스레드 할당 기법

김민철,이광엽 사단법인 인문사회과학기술융합학회 2017 예술인문사회융합멀티미디어논문지 Vol.7 No.10

CNN (Convolution neural network), which is used for image classification and speech recognition among neural networks learning based on positive data, has been continuously developed to have a high performance structure to date. There are many difficulties to utilize in an embedded system with limited resources. Therefore, we use GPU (General-Purpose Computing on Graphics Processing Units), which is used for general-purpose operation of GPU to solve the problem because we use pre-learned weights but there are still limitations. Since CNN performs simple and iterative operations, the computation speed varies greatly depending on the thread allocation and utilization method in the Single Instruction Multiple Thread (SIMT) based GPGPU. To solve this problem, there is a thread that needs to be relaxed when performing Convolution and Pooling operations with threads. The remaining threads have increased the operation speed by using the method used in the following feature maps and kernel calculations. 많은 양의 데이터 기반으로 학습하는 neural network 중 이미지 분류나 음성 인식 등에 사용되어 지고 있는 CNN(Convolution neural network)는 현재까지도 우수한 성능을 가진 구조로 계속적으로 발전되고 있다. 제한된 자원을 가진 임베디드 시스템에서 활용하기에는 많은 어려움이 있다. 그래서 미리 학습된 가중치를 사용하지만 여전히 한계점이 있기 때문에 이를 해결하기 위해 GPU의 범용 연산을 위해서 사용하는 GP-GPU(General-Purpose computing on Graphics Processing Units)를 활용하는 추세다. CNN은 단순하고 반복적인 연산을 수행하기 때문에 SIMT(Single Instruction Multiple Thread)기반의 GPGPU에서 스레드 할당과 활용 방법에 따라 연산 속도가 많이 달라진다. 스레드로 Convolution 연산과 Pooling 연산을 수행할 때 쉬어야 하는 스레드가 발생하는 데 이러한 문제를 해결하기 위해 남은 스레드가 다음 피쳐맵과 커널 계산에 활용되는 방법을 사용함으로써 연산 속도를 증가시켰다.

합성곱 신경망에 의한 과도 하중 감지에 대한 연구

강정호 한국기계가공학회 2024 한국기계가공학회지 Vol.23 No.1

Accurate load detection is important for the safe operation of mechanical structures. This study attempts toincorporate convolutional neural network technology to improve the accuracy of load detection. To increase thedetection accuracy of the excessive external load applied to the mechanical structure, it is necessary to check theamount of learning data along with the application of the Convolutional Neural Network technique of the artificialneural network. In this study, the amount of analysis data of the mechanical structure was insufficient forConvolutional Neural Network learning; therefore, the application of the stacked auto code technique was examinedas a method of securing the amount of data required for Convolutional Neural Network learning. In addition,appropriate conditions were suggested by researching the procedure of the Convolutional Neural Network for loaddetection and conditions appropriate for each function and variable. The accuracy was verified by detecting thehoisting load of the crawler crane from the load of the roller, which was the lower structure. The results of thisstudy can be applied to the analysis and design of various mechanical structures, and the possibility of using aConvolutional Neural Network to ensure machine safety and prevent accidents was verified.

Comparative Study of Ship Image Classification using Feedforward Neural Network and Convolutional Neural Network

Dae-Ki Kang The Institute of Internet 2024 International Journal of Internet, Broadcasting an Vol.16 No.3

In autonomous navigation systems, the need for fast and accurate image processing using deep learning and advanced sensor technologies is paramount. These systems rely heavily on the ability to process and interpret visual data swiftly and precisely to ensure safe and efficient navigation. Despite the critical importance of such capabilities, there has been a noticeable lack of research specifically focused on ship image classification for maritime applications. This gap highlights the necessity for more in-depth studies in this domain. In this paper, we aim to address this gap by presenting a comprehensive comparative study of ship image classification using two distinct neural network models: the Feedforward Neural Network (FNN) and the Convolutional Neural Network (CNN). Our study involves the application of both models to the task of classifying ship images, utilizing a dataset specifically prepared for this purpose. Through our analysis, we found that the Convolutional Neural Network demonstrates significantly more effective performance in accurately classifying ship images compared to the Feedforward Neural Network. The findings from this research are significant as they can contribute to the advancement of core source technologies for maritime autonomous navigation systems. By leveraging the superior image classification capabilities of convolutional neural networks, we can enhance the accuracy and reliability of these systems. This improvement is crucial for the development of more efficient and safer autonomous maritime operations, ultimately contributing to the broader field of autonomous transportation technology.

합성곱신경망을 이용한 타이타닉호의 재난 데이터 분석

이석준,심동희 차세대컨버전스정보서비스학회 2021 차세대컨버전스정보서비스기술논문지 Vol.10 No.1

Titanic disaster data is analyzed using the convolutional neural networks in this paper. Although many researchers have analyzed this data using the aritificial intelligence techniques such as k-nearest neighbors algorithm, Naive Bayes algorithm, decision tree algorithm, SVM algorithm and logistic regression algorithm. Although the artificial neural networks are also used in another researches, the convolutional neural networks model has not been used in the analysis for this titanic disaster data. LeNet-5 among the various convolutional neural networks model is used with Keras for the train in this analysis. The performance result is compared with the result of analysis using the mutlilayer neural networks uploaded in Kaggle web site. In the comparison of the number of parameters to be trained, LeNet-5 shows efficient than multilayer neural networks. In the results of test after trained process, LeNet-5 shows 0.856 in accuracy and 0.923 in precision. These performance measures are better than those of multilayer neural networks whose accuracy was 0.841 and the precision was 0.872. 본 연구에서는 합성곱신경망을 이용해서 타이타닉 재난데이터를 분석하였다. 그 동안 타이타닉 재난 데이터에 대한 k-최근접이웃 알고리즘, Naive Bayes 알고리즘, 의사결정트리 알고리즘 , SVM 알고리즘, 로지스틱 회귀 알고리즘 등 인공지능기법을 이용한 많은 분석이 이루어져왔다. 또한 인공신경망을 이용한 분석도 이루어져왔지만 합성곱신경망을 이용해서 이 타이타닉 재난데이터에 대한 분석은 없었다. 이 타이타닉 재난데이터 분석을 위하여 Keras를 이용하여 합성곱신경망중 LeNet-5를 사용해 모델링하여 훈련을 진행하였다. 이 성능결과를 Kaggle 웹사이트에 올려진 다중신경망을 이용한 분석결과와 비교하였다. LeNet-5를 이용한 경우 모델훈련에서 요구되는 훈련파라미터 개수의 비교에서 Kaggle 웹사이트에 올려진 다중신경망보다 월등히 적었다. 또한 훈련된 모델을 이용한 테스트 결과에서도 정확도는 0.856, 정밀도는 0.924를 나타냈다. 이 성능측정치는 Kaggle에 올려진 다중신경망의 정확도 0.842, 정밀도 0.873보다 우수한 것이다.

A Video Expression Recognition Method Based on Multi-mode Convolution Neural Network and Multiplicative Feature Fusion

( Qun Ren ) 한국정보처리학회 2021 Journal of information processing systems Vol.17 No.3

The existing video expression recognition methods mainly focus on the spatial feature extraction of video expression images, but tend to ignore the dynamic features of video sequences. To solve this problem, a multimode convolution neural network method is proposed to effectively improve the performance of facial expression recognition in video. Firstly, OpenFace 2.0 is used to detect face images in video, and two deep convolution neural networks are used to extract spatiotemporal expression features. Furthermore, spatial convolution neural network is used to extract the spatial information features of each static expression image, and the dynamic information feature is extracted from the optical flow information of multiple expression images based on temporal convolution neural network. Then, the spatiotemporal features learned by the two deep convolution neural networks are fused by multiplication. Finally, the fused features are input into support vector machine to realize the facial expression classification. Experimental results show that the recognition accuracy of the proposed method can reach 64.57% and 60.89%, respectively on RML and Baum-ls datasets. It is better than that of other contrast methods.

관심 문자열 인식 기술을 이용한 가스계량기 자동 검침 시스템

이교혁(Kyohyuk Lee),김태연(Taeyeon Kim),김우주(Wooju Kim) 한국지능정보시스템학회 2020 지능정보연구 Vol.26 No.2

In this paper, we suggest an application system architecture which provides accurate, fast and efficient automatic gasometer reading function. The system captures gasometer image using mobile device camera, transmits the image to a cloud server on top of private LTE network, and analyzes the image to extract character information of device ID and gas usage amount by selective optical character recognition based on deep learning technology. In general, there are many types of character in an image and optical character recognition technology extracts all character information in an image. But some applications need to ignore non-of-interest types of character and only have to focus on some specific types of characters. For an example of the application, automatic gasometer reading system only need to extract device ID and gas usage amount character information from gasometer images to send bill to users. Non-of-interest character strings, such as device type, manufacturer, manufacturing date, specification and etc., are not valuable information to the application. Thus, the application have to analyze point of interest region and specific types of characters to extract valuable information only. We adopted CNN (Convolutional Neural Network) based object detection and CRNN (Convolutional Recurrent Neural Network) technology for selective optical character recognition which only analyze point of interest region for selective character information extraction. We build up 3 neural networks for the application system. The first is a convolutional neural network which detects point of interest region of gas usage amount and device ID information character strings, the second is another convolutional neural network which transforms spatial information of point of interest region to spatial sequential feature vectors, and the third is bi-directional long short term memory network which converts spatial sequential information to character strings using time-series analysis mapping from feature vectors to character strings. In this research, point of interest character strings are device ID and gas usage amount. Device ID consists of 12 arabic character strings and gas usage amount consists of 4 ~ 5 arabic character strings. All system components are implemented in Amazon Web Service Cloud with Intel Zeon E5-2686 v4 CPU and NVidia TESLA V100 GPU. The system architecture adopts master-lave processing structure for efficient and fast parallel processing coping with about 700,000 requests per day. Mobile device captures gasometer image and transmits to master process in AWS cloud. Master process runs on Intel Zeon CPU and pushes reading request from mobile device to an input queue with FIFO (First In First Out) structure. Slave process consists of 3 types of deep neural networks which conduct character recognition process and runs on NVidia GPU module. Slave process is always polling the input queue to get recognition request. If there are some requests from master process in the input queue, slave process converts the image in the input queue to device ID character string, gas usage amount character string and position information of the strings, returns the information to output queue, and switch to idle mode to poll the input queue. Master process gets final information form the output queue and delivers the information to the mobile device. We used total 27,120 gasometer images for training, validation and testing of 3 types of deep neural network. 22,985 images were used for training and validation, 4,135 images were used for testing. We randomly splitted 22,985 images with 8:2 ratio for training and validation respectively for each training epoch. 4,135 test image were categorized into 5 types (Normal, noise, reflex, scale and slant). Normal data is clean image data, noise means image with noise signal, relfex means image with light reflection in gasometer region, scale means images with small object size due to long-distance capturin

Convolutional Neural Network Based Multi-feature Fusion for Non-rigid 3D Model Retrieval

( Hui Zeng ),( Yanrong Liu ),( Siqi Li ),( Jianyong Che ),( Xiuqing Wang ) 한국정보처리학회 2018 Journal of information processing systems Vol.14 No.1

This paper presents a novel convolutional neural network based multi-feature fusion learning method for nonrigid 3D model retrieval, which can investigate the useful discriminative information of the heat kernel signature (HKS) descriptor and the wave kernel signature (WKS) descriptor. At first, we compute the 2D shape distributions of the two kinds of descriptors to represent the 3D model and use them as the input to the networks. Then we construct two convolutional neural networks for the HKS distribution and the WKS distribution separately, and use the multi-feature fusion layer to connect them. The fusion layer not only can exploit more discriminative characteristics of the two descriptors, but also can complement the correlated information between the two kinds of descriptors. Furthermore, to further improve the performance of the description ability, the cross-connected layer is built to combine the low-level features with high-level features. Extensive experiments have validated the effectiveness of the designed multi-feature fusion learning method.