• Smart Media Journal
• /
• v.11 no.7
• /
• pp.94-103
• /
• 2022

Weeds bring disadvantages to crops since they can damage them, and a clean treatment with less pollution and contamination should be developed. Artificial intelligence gives new hope to agriculture to achieve smart farming. This study delivers an automated weeds growth point detection using deep learning. This study proposes a combination of semantic graphics for generating data annotation and U-Net with pre-trained deep learning as a backbone for locating the growth point of the weeds on the given field scene. The dataset was collected from an actual field. We measured the intersection over union, f1-score, precision, and recall to evaluate our method. Moreover, Mobilenet V2 was chosen as the backbone and compared with Resnet 34. The results showed that the proposed method was accurate enough to detect the growth point and handle the brightness variation. The best performance was achieved by Mobilenet V2 as a backbone with IoU 96.81%, precision 97.77%, recall 98.97%, and f1-score 97.30%.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.13 no.5
• /
• pp.1065-1070
• /
• 2018

These days there are many researches on the vision based technology using deep learning. The lots of studies on the intelligent operation and maintenance for railway station system used technologies with vision analysis function. This paper analyzes the papers which studied the intelligent station system with vision analysis function for passengers and facilities monitoring, platform monitoring, fire monitoring, and effective operation and design. Also, this paper proposes research which uses the more powerful vision technology with deep-learning for smart railway station system.

• Journal of the Korean Applied Science and Technology
• /
• v.37 no.3
• /
• pp.473-483
• /
• 2020

This paper proposes a tourism recommendation system in intelligent cloud environment using information of tourist behavior applied with perceived value. This proposed system applied tourist information and empirical analysis information that reflected the perceptual value of tourists in their behavior to the tourism recommendation system using wide and deep learning technology. This proposal system was applied to the tourism recommendation system by collecting and analyzing various tourist information that can be collected and analyzing the values that tourists were usually aware of and the intentions of people's behavior. It provides empirical information by analyzing and mapping the association of tourism information, perceived value and behavior to tourism platforms in various fields that have been used. In addition, the tourism recommendation system using wide and deep learning technology, which can achieve both memorization and generalization in one model by learning linear model components and neural only components together, and the method of pipeline operation was presented. As a result of applying wide and deep learning model, the recommendation system presented in this paper showed that the app subscription rate on the visiting page of the tourism-related app store increased by 3.9% compared to the control group, and the other 1% group applied a model using only the same variables and only the deep side of the neural network structure, resulting in a 1% increase in subscription rate compared to the model using only the deep side. In addition, by measuring the area (AUC) below the receiver operating characteristic curve for the dataset, offline AUC was also derived that the wide-and-deep learning model was somewhat higher, but more influential in online traffic.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.6
• /
• pp.1530-1544
• /
• 2023

Data monitoring is an important foundation of modern science. In most cases, the monitoring data is time-series data, which has high application value. The deep learning algorithm has a strong nonlinear fitting capability, which enables the recognition of time series by capturing anomalous information in time series. At present, the research of time series recognition based on deep learning is especially important for data monitoring. Deep learning algorithms require a large amount of data for training. However, abnormal sample is a small sample in time series, which means the number of abnormal time series can seriously affect the accuracy of recognition algorithm because of class imbalance. In order to increase the number of abnormal sample, a data augmentation method called GANBATS (GAN-based Bi-LSTM and Attention for Time Series) is proposed. In GANBATS, Bi-LSTM is introduced to extract the timing features and then transfer features to the generator network of GANBATS.GANBATS also modifies the discriminator network by adding an attention mechanism to achieve global attention for time series. At the end of discriminator, GANBATS is adding averagepooling layer, which merges temporal features to boost the operational efficiency. In this paper, four time series datasets and five data augmentation algorithms are used for comparison experiments. The generated data are measured by PRD(Percent Root Mean Square Difference) and DTW(Dynamic Time Warping). The experimental results show that GANBATS reduces up to 26.22 in PRD metric and 9.45 in DTW metric. In addition, this paper uses different algorithms to reconstruct the datasets and compare them by classification accuracy. The classification accuracy is improved by 6.44%-12.96% on four time series datasets.

• International Journal of Computer Science & Network Security
• /
• v.24 no.4
• /
• pp.87-106
• /
• 2024

Vehicle-to-Home (V2H) and Home Centralized Photovoltaic (HCPV) systems can address various energy storage issues and enhance demand response programs. Renewable energy, such as solar energy and wind turbines, address the energy gap. However, no energy management system is currently available to regulate the uncertainty of renewable energy sources, electric vehicles, and appliance consumption within a smart microgrid. Therefore, this study investigated the impact of solar photovoltaic (PV) panels, electric vehicles, and Micro-Grid (MG) storage on maximum solar radiation hours. Several Deep Learning (DL) algorithms were applied to account for the uncertainty. Moreover, a Reinforcement Learning HCPV (RL-HCPV) algorithm was created for efficient real-time energy scheduling decisions. The proposed algorithm managed the energy demand between PV solar energy generation and vehicle energy storage. RL-HCPV was modeled according to several constraints to meet household electricity demands in sunny and cloudy weather. Simulations demonstrated how the proposed RL-HCPV system could efficiently handle the demand response and how V2H can help to smooth the appliance load profile and reduce power consumption costs with sustainable power generation. The results demonstrated the advantages of utilizing RL and V2H as potential storage technology for smart buildings.

• Journal of Korea Multimedia Society
• /
• v.21 no.2
• /
• pp.188-198
• /
• 2018

With recent development of high-speed wide-area wireless networks and wide spread of highperformance wireless devices, the demand on seamless video streaming services in Long Term Evolution (LTE) network environments is ever increasing. To meet the demand and provide enhanced Quality of Experience (QoE) with mobile users, the Dynamic Adaptive Streaming over HTTP (DASH) has been actively studied to achieve QoE enhanced video streaming service in dynamic network environments. However, the existing DASH algorithm to select the quality of requesting video segments is based on a procedural algorithm so that it reveals a limitation to adapt its performance to dynamic network situations. To overcome this limitation this paper proposes a novel quality selection mechanism based on a Deep Q-Network (DQN) model, the DQN-based DASH ABR($DQN_{ABR}$) mechanism. The $DQN_{ABR}$ mechanism replaces the existing DASH ABR algorithm with an intelligent deep learning model which optimizes service quality to mobile users through reinforcement learning. Compared to the existing approaches, the experimental analysis shows that the proposed solution outperforms in terms of adapting to dynamic wireless network situations and improving QoE experience of end users.

• Korean Journal of Artificial Intelligence
• /
• v.5 no.2
• /
• pp.43-53
• /
• 2017

In this paper, the pattern of extracting the same expression is proposed by using the Android intelligent device to identify the facial expression. The understanding and expression of expression are very important to human computer interaction, and the technology to identify human expressions is very popular. Instead of searching for the emoticons that users often use, you can identify facial expressions with acamera, which is a useful technique that can be used now. This thesis puts forward the technology of the third data is available on the website of the set, use the content to improve the infrastructure of the facial expression recognition accuracy, in order to improve the synthesis of neural network algorithm, making the facial expression recognition model, the user's facial expressions and similar e xpressions, reached 66%.It doesn't need to search for emoticons. If you use the camera to recognize the expression, itwill appear emoticons immediately. So this service is the emoticons used when people send messages to others, and it can feel a lot of convenience. In countless emoticons, there is no need to find emoticons, which is an increasing trend in deep learning. So we need to use more suitable algorithm for expression recognition, and then improve accuracy.

• Journal of the Society of Naval Architects of Korea
• /
• v.59 no.2
• /
• pp.72-79
• /
• 2022

Deep-learning-based Super-Resolution (SR) methods were evaluated to reconstruct pressure fields with a high resolution from low-resolution images taken from a coarse grid simulation. In addition to a canonical SRCNN(super-resolution convolutional neural network) model, two modified models from SRCNN, adding an activation function (ReLU or Sigmoid function) to the output layer, were considered in the present study. High resolution images obtained by three models were more vivid and reliable qualitatively, compared with a conventional super-resolution method of bicubic interpolation. A quantitative comparison of statistical similarity showed that SRCNN model with Sigmoid function achieved best performance with less dependency on original resolution of input images.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.7
• /
• pp.2496-2512
• /
• 2021

The requirements for powerful computing capability, high capacity, low latency and low energy consumption of emerging services, pose severe challenges to the fifth-generation (5G) network. As a promising paradigm, mobile edge networks can provide services in proximity to users by deploying computing components and cache at the edge, which can effectively decrease service delay. However, the coexistence of heterogeneous services and the sharing of limited resources lead to the competition between various services for multiple resources. This paper considers two typical heterogeneous services: computing services and content delivery services, in order to properly configure resources, it is crucial to develop an effective offloading and caching strategies. Considering the high energy consumption of 5G base stations, this paper considers the hybrid energy supply model of traditional power grid and green energy. Therefore, it is necessary to design a reasonable association mechanism which can allocate more service load to base stations rich in green energy to improve the utilization of green energy. This paper formed the joint optimization problem of computing offloading, caching and resource allocation for heterogeneous services with the objective of minimizing the on-grid power consumption under the constraints of limited resources and QoS guarantee. Since the joint optimization problem is a mixed integer nonlinear programming problem that is impossible to solve, this paper uses deep reinforcement learning method to learn the optimal strategy through a lot of training. Extensive simulation experiments show that compared with other schemes, the proposed scheme can allocate resources to heterogeneous service according to the green energy distribution which can effectively reduce the traditional energy consumption.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.2
• /
• pp.751-770
• /
• 2019

Action recognition is an essential task in computer vision due to the variety of prospective applications, such as security surveillance, machine learning, and human-computer interaction. The availability of more video data than ever before and the lofty performance of deep convolutional neural networks also make it essential for action recognition in video. Unfortunately, limited crafted video features and the scarcity of benchmark datasets make it challenging to address the multi-person action recognition task in video data. In this work, we propose a deep convolutional neural network-based Effective Hybrid Learning (EHL) framework for two-person interaction classification in video data. Our approach exploits a pre-trained network model (the VGG16 from the University of Oxford Visual Geometry Group) and extends the Faster R-CNN (region-based convolutional neural network a state-of-the-art detector for image classification). We broaden a semi-supervised learning method combined with an active learning method to improve overall performance. Numerous types of two-person interactions exist in the real world, which makes this a challenging task. In our experiment, we consider a limited number of actions, such as hugging, fighting, linking arms, talking, and kidnapping in two environment such simple and complex. We show that our trained model with an active semi-supervised learning architecture gradually improves the performance. In a simple environment using an Intelligent Technology Laboratory (ITLab) dataset from Inha University, performance increased to 95.6% accuracy, and in a complex environment, performance reached 81% accuracy. Our method reduces data-labeling time, compared to supervised learning methods, for the ITLab dataset. We also conduct extensive experiment on Human Action Recognition benchmarks such as UT-Interaction dataset, HMDB51 dataset and obtain better performance than state-of-the-art approaches.