• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.105-116
• /
• 2022

Stay cables play an essential role in cable-stayed bridges. Severe vibrations and/or harsh environment may result in cable failures. Therefore, an efficient structural health monitoring (SHM) solution for cable damage detection is necessary. This study proposes a data-driven method for immediately detecting cable damage from measured cable forces by recognizing pattern transition from the intact condition when damage occurs. In the proposed method, pattern recognition for cable damage detection is realized by time series classification (TSC) using a deep learning (DL) model, namely, the long short term memory fully convolutional network (LSTM-FCN). First, a TSC classifier is trained and validated using the cable forces (or cable force ratios) collected from intact stay cables, setting the segmented data series as input and the cable (or cable pair) ID as class labels. Subsequently, the classifier is tested using the data collected under possible damaged conditions. Finally, the cable or cable pair corresponding to the least classification accuracy is recommended as the most probable damaged cable or cable pair. A case study using measured cable forces from an in-service cable-stayed bridge shows that the cable with damage can be correctly identified using the proposed DL-TSC method. Compared with existing cable damage detection methods in the literature, the DL-TSC method requires minor data preprocessing and feature engineering and thus enables fast and convenient early detection in real applications.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.3
• /
• pp.221-225
• /
• 2023

In this paper, we proposed an efficient method for rapid diagnosis of COVID-19 by voice. A novel Strided Convolution Separable Transformer (SC-SepTr) is proposed by modifying the conventional Separable Transformer (SepTr) for audio signal recognition. The proposed method reduces the memory and computational requirements to enable rapid diagnosis of COVID-19. As a result of experiments on Coswara, it was shown that the proposed method perform rapid diagnosis with guaranteeing Area Under the Curve (AUC) performance even for a relatively small amount of learning data.

• The KIPS Transactions:PartD
• /
• v.11D no.2
• /
• pp.443-452
• /
• 2004

The paper presents an intelligent space management system for ubiquitous computing. The system is basically a home/office automation system that could control light, electronic key, and home appliances such as TV and audio. On top of these basic capabilities, there are four elegant features in the system. First, we can access the system using either a cellular Phone or using a browser on the PC connected to the Internet, so that we control the system at any time and any place. Second, to provide more human-oriented interface, we integrate voice recognition functionalities into the system. Third, the system supports not only reactive services but also proactive services, based on the regularities of user behavior. Finally, by exploiting embedded technologies, the system could be run on the hardware that has less-processing power and storage. We have implemented the system on the embedded board consisting of StrongARM CPU with 205MHz, 32MB SDRAM, 16MB NOR-type flash memory, and Relay box. Under these hardware platforms, software components such as embedded Linux, HTK voice recognition tools, GoAhead Web Server, and GPIO driver are cooperated to support user-friendly intelligent space.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.9
• /
• pp.11-19
• /
• 2009

A RAM-based Neural Net(RBNN) which has multi-discriminators is more effective than RBNN with a discriminator. Experience Sensitive Cumulative Neural Network and 3-D Neuro System(3DNS) that accumulate the features point improved the performance of BNN, which were enabled to train additional and repeated patterns and extract a generalized pattern. In recognition process of Neural Net with multi-discriminator, the selection of class was decided by the value of MRD which calculates the accumulated sum of each class. But they had a saturation problem of its memory cells caused by learning volume increment. Therefore, the decision of MRD has a low performance because recognition rate is decreased by saturation. In this paper, we propose the method which improve the MRD ability. The method consists of the optimum MRD and the matching ratio prototype to generalized image, the cumulative filter ratio, the gap of prototype response MRD. We experimented the performance using NIST database of NIST without preprocessor, and compared this model with 3DNS. The proposed MRD method has more performance of recognition rate and more stable system for distortion of input pattern than 3DNS.

• Journal of Popular Narrative
• /
• v.26 no.1
• /
• pp.175-209
• /
• 2020

This paper examines how the cinematic representation of the Japanese military "comfort women" stimulates 'imagination' in the realm of everyday life and in the memory of the masses, creating a common awareness and affect. The history of the Japanese military "comfort women" was hidden for a long time, and it was not until the 1990s that it entered the field of public recognition. Such a transition can be attributed to the external and internal chronopolitics that made possible the testimony of the victims and the discourse of the "comfort women" issue. It shows the peculiar status of the comfort women history as 'politics of time'. In the same vein, the cinematic representations of the Japanese military "comfort women" can be found in similar chronopolitics. The 'comfort women' films have shown the dual time frame of the continuity and discontinuity of the 'silence'. In Korean film history, the chronotope of the reproduction of "comfort women" can be divided into four phases: 1) the fictional representations of "comfort women" before the 1990s 2) documentaries in the late 1990s as the work of testimony and history writing, 3) melodramatic transformation in the feature films in the 2000s, and 4) the diffusion of media and categories. The purpose of this article is to focus on the first phase and the third phase in which the issue of 'comfort women' is represented in the category of popular fiction films. While the "comfort women" representations before 1990 were strictly adhering to the framework of commercial movies and pursued the sexual exploitation of "comfort women" history, the recent films since the 2000s are experimenting with various attempts in the style of popular imagination. Especially, the emergence of 'comfort women' feature films in the 2000s, such as Spirit's Homecoming, I Can Speak, and Herstory, raise various questions as to whether we are "properly" aware of issues and how to remember and present the "cultural memory" of comfort women. Also, focusing on the cinematic representation strategies of the 2000s "comfort women", this article discusses the popular politics of melodrama, the representation of victims and violence, and the feature of 'comfort women' as meta-memory. As a melodramatic imagination and meta-memory for the historical trauma, the "comfort women" drama shows the historical, political, and aesthetic gateways to which the "comfort women" problem must pass. As we have seen in recent fiction films, the issue of "comfort women" goes beyond transnational relations between Korea and Japan; it demands a postcolonial task to dismantle the old colonial structure and explores a transnational project in which women's movements and human rights movements are linked internationally.

• 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.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

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 training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.1-25
• /
• 2020

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 capturing and slant means images which is not horizontally flat. Final character string recognition accuracies for device ID and gas usage amount of normal data are 0.960 and 0.864 respectively.

• Journal of Educational Research in Mathematics
• /
• v.25 no.4
• /
• pp.571-598
• /
• 2015

This study aimed to assess the spatial cognition strategies and roles taken by students in the process of solving spatial tasks. For the analysis, this study developed two spatial tests based on the mental rotation test, which were taken by 63 students in their final year in elementary schools. The results of this study showed that in terms of the method of approaching the tasks, students took the comprehensive approach and the partial approach. When solving the tasks, the students were shown to use the imagery thinking or analytic thinking method. In terms of perspective, the students rotated the object or change their perspectives. A comparison of the methods used by the students revealed that when approaching the tasks, the group of students who chose the partial approach had higher scores. In terms of solving the tasks the analytic thinking method, and in terms of perspective, changing perspectives were shown to be more effective. Such effective methods were used more frequently in discrimination tasks than in recognition tasks, and in more complicated items, than in less complicated items. In conclusion, the results of this study suggested that the partial, analytic approach and the change of perspectives are useful strategies in solving tasks which require high cognitive effort.

• Food Science of Animal Resources
• /
• v.40 no.6
• /
• pp.1014-1032
• /
• 2020

This work aimed to investigate the psychobiotic effects of six bacterial strains on the mind and behavior of male Wistar rats. The probiotic (PRO) group (n=7) were rats pre-treated with antibiotics for 7 days followed by 14-day probiotic administration, antibiotics (ANT) group (n=7) were rats treated with antibiotics for 21 days without probiotics. The control (CON) group (n=7) were rats that received sham treatment for 21 days. The six bacterial strains with probiotic properties were mostly isolated from Thai fermented foods; Pedicoccus pentosaceus WS11, Lactobacillus plantarum SK321, L. fermentum SK324, L. brevis TRBC 3003, Bifidobacterium adolescentis TBRC 7154 and Lactococcus lactis subsp. lactis TBRC 375. The probiotics were freeze-dried into powder (6×10⁹ CFU/5 g) and administered to the PRO group via oral gavage. Behavioral tests were performed. The PRO group displayed significantly reduced anxiety level and increased locomotor function using a marble burying test and open field test, respectively and significantly improved short-term memory performance using a novel object recognition test. Antibiotics significantly reduced microbial counts in rat feces in the ANT group by 100 fold compared to the PRO group. Probiotics significantly enhanced antioxidant enzymatic and non-enzymatic defenses in rat brains as assessed using catalase activity and ferric reducing antioxidant power assay, respectively. Probiotics also showed neuroprotective effects with less pyknotic cells and lower frequency of vacuolization in cerebral cortex. This multi-strain probiotic formulation from Thai fermented foods may offer a potential to develop psychobiotic-rich functional foods to modulate human mind and behaviors.