• Journal of the Korean Institute of Intelligent Systems
• /
• v.27 no.1
• /
• pp.22-28
• /
• 2017

The Convolutional Neural Network(CNN) can be divided into two stages: feature extraction and classification. The hyperparameters such as kernel size, number of channels, and stride in the feature extraction step affect the overall performance of CNN as well as determining the structure of CNN. In this paper, we propose a method to optimize the hyperparameter in CNN feature extraction stage using Parameter-Setting-Free Harmony Search (PSF-HS) algorithm. After setting the overall structure of CNN, hyperparameter was set as a variable and the hyperparameter was optimized by applying PSF-HS algorithm. The simulation was conducted using MATLAB, and CNN learned and tested using mnist data. We update the parameters for a total of 500 times, and it is confirmed that the structure with the highest accuracy among the CNN structures obtained by the proposed method classifies the mnist data with an accuracy of 99.28%.

• Journal of Broadcast Engineering
• /
• v.23 no.3
• /
• pp.369-382
• /
• 2018

We propose a single-shot high dynamic range (HDR) imaging algorithm using a deep convolutional neural network (CNN) for row-wise varying exposures in a single image. The proposed algorithm restores missing information resulting from under- and/or over-exposed pixels in an input image and reconstructs the raw radiance map. The main contribution of this work is the development of a loss function for the CNN employing the human visual system (HVS) properties. Then, the HDR image is obtained by applying a demosaicing algorithm. Experimental results demonstrate that the proposed algorithm provides higher-quality HDR images than conventional algorithms.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.120-120
• /
• 2020

Spatial precipitation data is one of the essential components in modeling hydrological problems. The estimation of these data has achieved significant achievements own to the recent advances in remote sensing technology. However, there are still gaps between the satellite-derived rainfall data and observed data due to the significant dependence of rainfall on spatial and temporal characteristics. An effective approach based on the Convolutional Neural Network (CNN) model to correct the satellite-derived rainfall data is proposed in this study. The Mekong River basin, one of the largest river system in the world, was selected as a case study. The two gridded precipitation data sets with a spatial resolution of 0.25 degrees used in the CNN model are APHRODITE (Asian Precipitation - Highly-Resolved Observational Data Integration Towards Evaluation) and PERSIANN-CDR (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks). In particular, PERSIANN-CDR data is exploited as satellite-based precipitation data and APHRODITE data is considered as observed rainfall data. In addition to developing a CNN model to correct the satellite-based rain data, another statistical method based on standard deviations for precipitation bias correction was also mentioned in this study. Estimated results indicate that the CNN model illustrates better performance both in spatial and temporal correlation when compared to the standard deviation method. The finding of this study indicated that the CNN model could produce reliable estimates for the gridded precipitation bias correction problem.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.5
• /
• pp.14-21
• /
• 2020

In order to provide confidential services between two communicating parties, block data encryption using a symmetric secret key is applied. A power analysis attack on a cryptosystem is a side channel-analysis method that can extract a secret key by measuring the power consumption traces of the crypto device. In this paper, we propose an attack model that can recover the secret key using a power analysis attack based on a deep learning convolutional neural network (CNN) algorithm. Considering that the CNN algorithm is suitable for image analysis, we particularly adopt the recurrence plot (RP) signal processing method, which transforms the one-dimensional power trace into two-dimensional data. As a result of executing the proposed CNN attack model on an XMEGA128 experimental board that implemented the AES-128 encryption algorithm, we recovered the secret key with 22.23% accuracy using raw power consumption traces, and obtained 97.93% accuracy using power traces on which we applied the RP processing method.

• Journal of Korea Multimedia Society
• /
• v.25 no.11
• /
• pp.1643-1652
• /
• 2022

Most of the existing machine reading research has used Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN) algorithms as networks. Among them, RNN was slow in training, and Question Answering Network (QANet) was announced to improve training speed. QANet is a model composed of CNN and self-attention. CNN extracts semantic and syntactic information well from the local corpus, but there is a limit to extracting the corresponding information from the global corpus. Graph Convolutional Networks (GCN) extracts semantic and syntactic information relatively well from the global corpus. In this paper, to take advantage of this strength of GCN, we propose I-QANet, which changed the CNN of QANet to GCN. The proposed model performed 1.2 times faster than the baseline in the Stanford Question Answering Dataset (SQuAD) dataset and showed 0.2% higher performance in Exact Match (EM) and 0.7% higher in F1. Furthermore, in the Korean Question Answering Dataset (KorQuAD) dataset consisting only of Korean, the learning time was 1.1 times faster than the baseline, and the EM and F1 performance were also 0.9% and 0.7% higher, respectively.

• Nuclear Engineering and Technology
• /
• v.54 no.12
• /
• pp.4751-4758
• /
• 2022

Two-phase flow may almost exist in every branch of the energy industry. For the corresponding engineering design, it is very essential and crucial to monitor flow patterns and their transitions accurately. With the high-speed development and success of deep learning based on convolutional neural network (CNN), the study of flow pattern identification recently almost focused on this methodology. Additionally, the photographing technique has attractive implementation features as well, since it is normally considerably less expensive than other techniques. The development of such a two-phase flow pattern online monitoring system is the objective of this work, which seldom studied before. The ongoing preliminary engineering design (including hardware and software) of the system are introduced. The flow pattern identification method based on CNNs and transfer learning was discussed in detail. Several potential CNN candidates such as ALexNet, VggNet16 and ResNets were introduced and compared with each other based on a flow pattern dataset. According to the results, ResNet50 is the most promising CNN network for the system owing to its high precision, fast classification and strong robustness. This work can be a reference for the online monitoring system design in the energy system.

• Journal of Information Science Theory and Practice
• /
• v.7 no.2
• /
• pp.32-39
• /
• 2019

A convolutional neural network (CNN) has been widely used in facial expression recognition (FER) because it can automatically learn discriminative appearance features from an expression image. To make full use of its discriminating capability, this paper suggests a simple but effective method for CNN based FER. Specifically, instead of an original expression image that contains facial appearance only, the expression image with facial geometry visualization is used as input to CNN. In this way, geometric and appearance features could be simultaneously learned, making CNN more discriminative for FER. A simple CNN extension is also presented in this paper, aiming to utilize geometric expression change derived from an expression image sequence. Experimental results on two public datasets (CK+ and MMI) show that CNN using facial geometry visualization clearly outperforms the conventional CNN using facial appearance only.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.370-373
• /
• 2022

Recently, various deep learning framework models such as Tensorflow, Pytorch, Keras, etc. have appeared. In addition, CNN (Convolutional Neural Network) is applied to image recognition using frameworks such as Tensorflow, Pytorch, and Keras, and the optimization model in image classification is mainly used. In this paper, based on the results of training the CNN model with the Paitotchi and tensor flow frameworks most often used in the field of deep learning image recognition, the two frameworks are compared and analyzed for image analysis. Derived an optimized framework.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.12 no.2
• /
• pp.113-120
• /
• 2017

In this paper, we present a night-time vehicle detection method using CNN (Convolutional Neural Network) classification. The camera based night-time vehicle detection plays an important role on various advanced driver assistance systems (ADAS) such as automatic head-lamp control system. The method consists mainly of thresholding, labeling and classification steps. The classification step is implemented by existing CIFAR-10 model CNN. Through the simulations tested on real road video, we show that CNN classification is a good alternative for night-time vehicle detection.

• Phonetics and Speech Sciences
• /
• v.10 no.1
• /
• pp.33-38
• /
• 2018

Although deep neural network-based acoustic models have greatly improved the performance of automatic speech recognition (ASR), reverberation still degrades the performance of distant speech recognition in indoor environments. In this paper, we adopt the DenseNet, which has shown great performance results in image classification tasks, to improve the performance of reverberant speech recognition. The DenseNet enables the deep convolutional neural network (CNN) to be effectively trained by concatenating feature maps in each convolutional layer. In addition, we extend the concept of multi-resolution CNN to multi-resolution DenseNet for robust speech recognition in reverberant environments. We evaluate the performance of reverberant speech recognition on the single-channel ASR task in reverberant voice enhancement and recognition benchmark (REVERB) challenge 2014. According to the experimental results, the DenseNet-based acoustic models show better performance than do the conventional CNN-based ones, and the multi-resolution DenseNet provides additional performance improvement.