• Journal of KIISE
• /
• v.44 no.6
• /
• pp.607-612
• /
• 2017

This paper presents an attempt to apply Deep Learning technology to solve the problem of forecasting floods in urban areas. We employ Recurrent Neural Networks (RNNs), which are suitable for analyzing time series data, to learn observed data of river water and to predict the water level. To test the model, we use water observation data of a station in the Trinity river, Texas, the U.S., with data from 2013 to 2015 for training and data in 2016 for testing. Input of the neural networks is a 16-record-length sequence of 15-minute-interval time-series data, and output is the predicted value of the water level at the next 30 minutes and 60 minutes. In the experiment, we compare three Deep Learning models including standard RNN, RNN trained with Back Propagation Through Time (RNN-BPTT), and Long Short-Term Memory (LSTM). The prediction quality of LSTM can obtain Nash Efficiency exceeding 0.98, while the standard RNN and RNN-BPTT also provide very high accuracy.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.1
• /
• pp.146-151
• /
• 2021

Currently, the demand for vehicles from one family is increasing in many countries at home and abroad, reducing the number of people on the vehicle and increasing the number of vehicles on the road. The multi-passenger lane system, which is available to solve the problem of traffic congestion, is being implemented. The system allows police to monitor fast-moving vehicles with their own eyes to crack down on illegal vehicles, which is less accurate and accompanied by the risk of accidents. To address these problems, applying deep learning object recognition techniques using images from road sites will solve the aforementioned problems. Therefore, in this paper, we compare and analyze the performance of existing deep learning models, select a deep learning model that can identify real-time vehicle occupants through video, and propose a vehicle occupancy detection algorithm that complements the object-ident model's problems.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.4
• /
• pp.611-616
• /
• 2023

ECG(electrocardiogram) is a test used to measure the rate and regularity of heartbeats, as well as the size and position of the chambers, the presence of any damage to the heart, and the cause of all heart diseases can be found. Because the ECG signal obtained using the ECG-KIT includes noise in the ECG signal, noise must be removed from the ECG signal to apply to the deep learning. In this paper, the noise of the ECG signal was removed using the digital IIR Butterworth low-pass filter. When the performance evaluation of the three activation functions, sigmoid(), ReLU(), and tanh() functions, was compared using the deep learning model of LSTM, it was confirmed that the activation function with the smallest error was the tanh() function. Also, When the performance evaluation and elapsed time were compared for LSTM and GRU models, it was confirmed that the GRU model was superior to the LSTM model.

• Korean Journal of Remote Sensing
• /
• v.39 no.2
• /
• pp.183-192
• /
• 2023

The remotely sensed data, such as satellite imagery and aerial photos, can be used to extract and detect some objects in the image through image interpretation and processing techniques. Significantly, the possibility for utilizing digital map updating and land monitoring has been increased through automatic object detection since spatial resolution of remotely sensed data has improved and technologies about deep learning have been developed. In this paper, we tried to extract plastic greenhouses into aerial orthophotos by using fully convolutional densely connected convolutional network (FC-DenseNet), one of the representative deep learning models for semantic segmentation. Then, a quantitative analysis of extraction results had performed. Using the farm map of the Ministry of Agriculture, Food and Rural Affairsin Korea, training data was generated by labeling plastic greenhouses into Damyang and Miryang areas. And then, FC-DenseNet was trained through a training dataset. To apply the deep learning model in the remotely sensed imagery, instance norm, which can maintain the spectral characteristics of bands, was used as normalization. In addition, optimal weights for each band were determined by adding attention modules in the deep learning model. In the experiments, it was found that a deep learning model can extract plastic greenhouses. These results can be applied to digital map updating of Farm-map and landcover maps.

• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1407-1416
• /
• 2018

The purpose of this study was to train a model for the ulna and radius bone segmentation based on Convolutional Neural Networks and to verify the segmentation model. The data consisted of 840 training data, 210 tuning data, and 200 verification data. The learning model for the ulna and radius bone bwas based on U-Net (19 convolutional and 8 maximum pooling) and trained with 8 batch sizes, 0.0001 learning rate, and 200 epochs. As a result, the average sensitivity of the training data was 0.998, the specificity was 0.972, the accuracy was 0.979, and the Dice's similarity coefficient was 0.968. In the validation data, the average sensitivity was 0.961, specificity was 0.978, accuracy was 0.972, and Dice's similarity coefficient was 0.961. The performance of deep convolutional neural network based models for the segmentation was good for ulna and radius bone.

• Journal of the Korean Society of Industry Convergence
• /
• v.23 no.6_2
• /
• pp.979-987
• /
• 2020

This study used the Deep Learning models used in previous studies, we selected the basic model. The selected model was selected as ZFNet among ZFNet, Googlenet and ResNet, and the object was detected using a ZFNet based FRCNN. In order to reduce the detection error rate of FRCNN, location of four types of objects detected inside the image was designed by SVM classifier and location-based filtering was applied. As simulation results, it showed similar performance to the lane marking classification method with conventional 경계 detection, with an average accuracy of about 88.8%. In addition, studies using the Linear-parabolic Model showed a processing speed of 165.65ms with a minimum resolution of 600 × 800, but in this study, the resolution was treated at about 33ms with an input resolution image of 1280 × 960, so it was possible to classify lane marking at a faster rate than the previous study by CNN-based End to End method.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.1
• /
• pp.147-165
• /
• 2021

In this work, we empirically evaluated the efficiency of the recent EfficientNetB0 model to identify and diagnose malaria parasite infections in blood smears. The dataset used was collected and classified by relevant experts from the Lister Hill National Centre for Biomedical Communications (LHNCBC). We prepared our samples with minimal image transformations as opposed to others, as we focused more on the feature extraction capability of the EfficientNetB0 baseline model. We applied transfer learning to increase the initial feature sets and reduced the training time to train our model. We then fine-tuned it to work with our proposed layers and re-trained the entire model to learn from our prepared dataset. The highest overall accuracy attained from our evaluated results was 94.70% from fifty epochs and followed by 94.68% within just ten. Additional visualization and analysis using the Gradient-weighted Class Activation Mapping (Grad-CAM) algorithm visualized how effectively our fine-tuned EfficientNetB0 detected infections better than other recent state-of-the-art DCNN models. This study, therefore, concludes that when fine-tuned, the recent EfficientNetB0 will generate highly accurate deep learning solutions for the identification of malaria parasites in blood smears without the need for stringent pre-processing, optimization, or data augmentation of images.

• International Journal of Internet, Broadcasting and Communication
• /
• v.14 no.1
• /
• pp.136-141
• /
• 2022

A very complex task in deep learning such as image classification must be solved with the help of neural networks and activation functions. The backpropagation algorithm advances backward from the output layer towards the input layer, the gradients often get smaller and smaller and approach zero which eventually leaves the weights of the initial or lower layers nearly unchanged, as a result, the gradient descent never converges to the optimum. We propose a two-factor non-saturating activation functions known as Bea-Mish for machine learning applications in deep neural networks. Our method uses two factors, beta (𝛽) and alpha (𝛼), to normalize the area below the boundary in the Mish activation function and we regard these elements as Bea. Bea-Mish provide a clear understanding of the behaviors and conditions governing this regularization term can lead to a more principled approach for constructing better performing activation functions. We evaluate Bea-Mish results against Mish and Swish activation functions in various models and data sets. Empirical results show that our approach (Bea-Mish) outperforms native Mish using SqueezeNet backbone with an average precision (AP50val) of 2.51% in CIFAR-10 and top-1accuracy in ResNet-50 on ImageNet-1k. shows an improvement of 1.20%.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.3
• /
• pp.487-499
• /
• 2022

Machine learning such as deep learning have been widely used in recent years. Recently deep learning is performed in a trusted execution environment such as ARM TrustZone to improve security in edge devices and embedded devices with low computing resource. To mitigate this problem, we propose TPMP that efficiently uses the limited memory of TEE through DNN model partitioning. TPMP achieves high confidentiality of DNN by performing DNN models that could not be run with existing memory scheduling methods in TEE through optimized memory scheduling. TPMP required a similar amount of computational resources to previous methodologies.

• Smart Structures and Systems
• /
• v.31 no.4
• /
• pp.383-392
• /
• 2023

Post-earthquake building condition assessment is crucial for subsequent rescue and remediation and can be automated by emerging computer vision and deep learning technologies. This study is based on an endeavour for the 2nd International Competition of Structural Health Monitoring (IC-SHM 2021). The task package includes five image segmentation objectives - defects (crack/spall/rebar exposure), structural component, and damage state. The structural component and damage state tasks are identified as the priority that can form actionable decisions. A multi-task Convolutional Neural Network (CNN) is proposed to conduct the two major tasks simultaneously. The rest 3 sub-tasks (spall/crack/rebar exposure) were incorporated as auxiliary tasks. By synchronously learning defect information (spall/crack/rebar exposure), the multi-task CNN model outperforms the counterpart single-task models in recognizing structural components and estimating damage states. Particularly, the pixel-level damage state estimation witnesses a mIoU (mean intersection over union) improvement from 0.5855 to 0.6374. For the defect detection tasks, rebar exposure is omitted due to the extremely biased sample distribution. The segmentations of crack and spall are automated by single-task U-Net but with extra efforts to resample the provided data. The segmentation of small objects (spall and crack) benefits from the resampling method, with a substantial IoU increment of nearly 10%.