• Journal of the Korean Society of Safety
• /
• v.35 no.2
• /
• pp.94-99
• /
• 2020

Bikers can be subjected to injuries from unexpected accidents even if they wear basic helmets. A properly designed airbag can efficiently protect the critical areas of the human body. This study introduces a wearable smart airbag system using machine learning techniques to protect human neck and shoulders. When a bicycle accident happens, a microprocessor analyzes the biker's motion data to recognize if it is a critical accident by comparing with accident classification models. These models are trained by a variety of possible accidents through machine learning techniques, like k-means and SVM methods. When the microprocessor decides it is a critical accident, it issues an actuation signal for the gas inflater to inflate the airbag. A protype of the wearable smart airbag with the machine learning techniques is developed and its performance is tested using a human dummy mounted on a moving cart.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.1
• /
• pp.617-623
• /
• 2024

Identifying emotions through speech poses a significant challenge due to the complex relationship between language and emotions. Our paper aims to take on this challenge by employing feature engineering to identify emotions in speech through a multimodal classification task involving both speech and text data. We evaluated two classifiers-Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM)-both integrated with a BERT-based pre-trained model. Our assessment covers various performance metrics (accuracy, F-score, precision, and recall) across different experimental setups). The findings highlight the impressive proficiency of two models in accurately discerning emotions from both text and speech data.

• The Journal of the Korea Contents Association
• /
• v.10 no.5
• /
• pp.115-123
• /
• 2010

Because of development of information technology, moving picture can run various platforms. We should consider and apply users' attitude as well as production technique because convergence between mobile and media technology may be increased full-browsing service using mobile device. Previous research related to production technique in various platforms only focus on video quality and adjustment of screen size. However, besides of technical side, production techniques should be changed such as image production as well as image editing by point of view aesthetic. Mise-en-scene such as camera angle, composition, and lighting is changed due to HD image. Also image production should be changed to a suitable full-browsing service using mobile device. Therefore, we would explore a new suitable production techniques and image editing for smart phone. To propose production techniques for smart phone, we used E-learning production system, which are transition, editing technique for suitable converting system. Such as new attempts are leading to new paradigm and establishing their position by applying characteries such as openness, timeliness to mobile. Also it can be extended individual area and established as expression and play tool.

• Tunnel and Underground Space
• /
• v.33 no.4
• /
• pp.209-227
• /
• 2023

The government continues to announce measures to revitalize smart construction technology based on BIM for productivity innovation in the construction industry. In the design phase, the goal is design automation and optimization by converging BIM Data and other advanced technologies. Accordingly, in the basic design of the Namhae Seomyeon-Yeosu Sindeok National Road Construction Project, a domestic undersea tunnel project, BIM-based design was carried out by developing tunnel design automation technology using 3D spatial information according to the tunnel design process. In order to derive the optimal alignment, more than 10,000 alignment cases were generated in 36hr using the generative design technique and a quantitative evaluation of the objective functions defined by the designer was performed. AI-based ground classification and 3D Geo Model were established to evaluate the economic feasibility and stability of the optimal alignment. AI-based ground classification has improved its precision by performing about 30 types of ground classification per borehole, and in the case of the 3D Geo Model, its utilization can be expected in that it can accumulate ground data added during construction. In the case of 3D blasting design, the optimal charge weight was derived in 5 minutes by reviewing all security objects on the project range on Dynamo, and the design result was visualized in 3D space for intuitive and convenient construction management so that it could be used directly during construction.

• Smart Media Journal
• /
• v.9 no.4
• /
• pp.26-35
• /
• 2020

Attempts to predict future stock prices have been studied steadily since the past. However, unlike general time-series data, financial time-series data has various obstacles to making predictions such as non-stationarity, long-term dependence, and non-linearity. In addition, variables of a wide range of data have limitations in the selection by humans, and the model should be able to automatically extract variables well. In this paper, we propose a 'sliding time step normalization' method that can normalize non-stationary data and LSTM autoencoder to compress variables from all variables. and 'moving transfer learning', which divides periods and performs transfer learning. In addition, the experiment shows that the performance is superior when using as many variables as possible through the neural network rather than using only 100 major financial variables and by using 'sliding time step normalization' to normalize the non-stationarity of data in all sections, it is shown to be effective in improving performance. 'moving transfer learning' shows that it is effective in improving the performance in long test intervals by evaluating the performance of the model and performing transfer learning in the test interval for each step.

• The Journal of the Convergence on Culture Technology
• /
• v.6 no.4
• /
• pp.807-812
• /
• 2020

In order to prevent the national wide spread of the COVID-19 virus, the government enforces to use an electronic access registration system for public facilities to effectively track and manage the spread. Initially, there was a lot of hassle to write a directory, but recently a system for creating an electronic access list using QR codes, what is called KI-Pass, is mainly used. However, the procedure for generating a QR code is somewhat cumbersome. In this paper, we propose a new electronic access registration system that does not require QR code. This system effectively controls the suspicious visitor by using a mask wearing discriminator which has been implemented using deep learning technology, and a non-contact thermometer package. In addition, by linking the beacon, a short-range wireless communication technology, and the visitor's smartphone application, basic information of the facility visitor is automatically registered to KDCA through the server. On the other hand, the user access information registered in the server is encrypted and stored, and is automatically destroyed after up to 4 weeks. This system is expected to be very effective in preventing the spread of other new infectious diseases as well as responding to the coronavirus which is recording a high spread worldwide.

• The Journal of the Korea Contents Association
• /
• v.21 no.6
• /
• pp.701-715
• /
• 2021

The drastic advance of recent deep learning technologies is heavily dependent on training datasets which are essential to train models by themselves with less human efforts. In comparison with the work to design deep learning models, preparing datasets is a long haul; at the moment, in the domain of vision intelligent, datasets are still being made by handwork requiring a lot of time and efforts, where workers need to directly make labels on each image usually with GUI-based labeling tools. In this paper, we overview the current status of vision datasets focusing on what datasets are being shared and how they are prepared with various labeling tools. Particularly, in order to relieve the repetitive and tiring labeling work, we present an interactive smart image annotating system with which the annotation work can be transformed from the direct human-only manual labeling to a correction-after-checking by means of a support of automatic labeling. In an experiment, we show that automatic labeling can greatly improve the productivity of datasets especially reducing time and efforts to specify regions of objects found in images. Finally, we discuss critical issues that we faced in the experiment to our annotation system and describe future work to raise the productivity of image datasets creation for accelerating AI technology.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1373-1380
• /
• 2019

In order to increase energy efficiency, developments of the advanced metering infrastructure (AMI) in the smart grid technology have recently been actively conducted. An essential part of AMI is analyzing power consumption and forecasting consumption patterns. In this paper, we analyze the power consumption and summarized the data errors. Monthly power consumption patterns are also analyzed using the k-means clustering algorithm. Forecasting the consumption pattern by each household is difficult. Therefore, we first classify the data into 100 clusters and then predict the average of the next day as the daily average of the clusters based on the deep neural network. Using practically collected AMI data, we analyzed the data errors and could successfully conducted power forecasting based on a clustering technique.

• Journal of the Institute of Convergence Signal Processing
• /
• v.21 no.1
• /
• pp.29-37
• /
• 2020

This paper implements a method of extracting an object about a dog through real-time image analysis and classifying dog behaviors from the extracted images. The Darknet YOLO was used to detect dog objects, and the Teachable Machine provided by Google was used to classify behavior patterns from the extracted images. The trained Teachable Machine is saved in Google Drive and can be used by ml5.js implemented on a node.js server. By implementing an interactive web server using a socket.io module on the node.js server, the classified results are transmitted to the user's smart phone or PC in real time so that it can be checked anytime, anywhere.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.