• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.30 no.6
• /
• pp.1067-1078
• /
• 2020

With the development of internet technology, a lot of content is produced, and the demand for it is increasing. Accordingly, the number of contents in circulation is increasing, while the number of distributing illegal copies that infringe on copyright is also increasing. The Korea Copyright Protection Agency operates a illegal content obstruction program based on substring matching, and it is difficult to accurately search because a large number of noises are inserted to bypass this. Recently, researches using natural language processing and AI deep learning technologies to remove noise and various blockchain technologies for copyright protection are being studied, but there are limitations. In this paper, noise is removed from data collected online, and keyword-based illegal copies are searched. In addition, the same heavy uploader is estimated through profiling analysis for heavy uploaders. In the future, it is expected that copyright damage will be minimized if the illegal copy search technology and blocking and response technology are combined based on the results of profiling analysis for heavy uploaders.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.24 no.2
• /
• pp.175-186
• /
• 2021

In the field of automatic target recognition(ATR) with synthetic aperture radar(SAR) imagery, it is usually impractical to obtain SAR target images covering a full range of aspect views. When the database consists of SAR target images with limited angular diversity, it can lead to performance degradation of the SAR-ATR system. To address this problem, this paper proposes a deep learning-based method where channel attention modules(CAMs) are inserted to a convolutional neural network(CNN). Motivated by the idea of the squeeze-and-excitation(SE) network, the CAM is considered to help improve recognition performance by selectively emphasizing discriminative features and suppressing ones with less information. After testing various CAM types included in the ResNet18-type base network, the SE CAM and its modified forms are applied to SAR target recognition using MSTAR dataset with different reduction ratios in order to validate recognition performance improvement under the limited angular diversity condition.

• Journal of IKEEE
• /
• v.24 no.4
• /
• pp.1050-1057
• /
• 2020

Immersive 360-degree media has a problem of slowing down the video recognition speed when the video is processed by the conventional method using a variety of rendering methods, and the video size becomes larger with higher quality and extra-large volume than the existing video. In addition, in most cases, only one scene is captured by fixing the camera in a specific place due to the characteristics of the immersive 360-degree media, it is not necessary to extract feature information from all scenes. In this paper, we propose a reference frame selection method for immersive 360-degree media and describe its application process to copyright protection technology. In the proposed method, three pre-processing processes such as frame extraction of immersive 360 media, frame downsizing, and spherical form rendering are performed. In the rendering process, the video is divided into 16 frames and captured. In the central part where there is much object information, an object is extracted using an RGB vector per pixel and deep learning, and a reference frame is selected using object feature information.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.9
• /
• pp.1224-1230
• /
• 2020

Since the spread of smart phones, interest in wearable devices has increased and diversified, and is closely related to the lives of users, and has been used as a method for providing personalized services. In this paper, we propose a method to detect the user's behavior by applying information from a 3-axis acceleration sensor and a 3-axis gyro sensor embedded in a smartphone to a convolutional neural network. Human behavior differs according to the size and range of motion, starting and ending time, including the duration of the signal data constituting the motion. Therefore, there is a performance problem for accuracy when applied to a convolutional neural network as it is. Therefore, we proposed a Time-Division Feature Fusion Convolutional Neural Network (TDFFCNN) that learns the characteristics of the sensor data segmented over time. The proposed method outperformed other classifiers such as SVM, IBk, convolutional neural network, and long-term memory circulatory neural network.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.1
• /
• pp.57-64
• /
• 2020

Recently, due to the proliferation of IoT sensors, the development of big data and artificial intelligence, time series prediction research on fine dust pollution is actively conducted. However, because the data representing fine dust contamination changes rapidly, traditional time series prediction methods do not provide a level of accuracy that can be used in the field. In this paper, we propose a method that reflects the classification results of environmental conditions through CNN when predicting micro dust contamination using LSTM. Although LSTM and CNN are independent, they are integrated into one network through the interface, so this method is easier to understand than the application LSTM. In the verification experiments of the proposed method using Beijing PM2.5 data, the prediction accuracy and predictive power for the timing of change were consistently improved in various experimental cases.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.4
• /
• pp.459-466
• /
• 2020

The sound collected through the multi-rotor unmanned aerial vehicle (UAV) includes the ego noise generated by the motor or propeller, or the wind noise generated during the flight, and thus the quality is greatly impaired. In a multi-rotor UAV environment, both the magnitude and phase of the target sound are greatly corrupted, so it is necessary to enhance the sound in consideration of both the magnitude and phase. However, it is difficult to improve the phase because it does not show the structural characteristics. in this study, we propose a CNN-based complex spectrogram enhancement method that removes noise based on complex spectrogram that can represent both magnitude and phase. Experimental results reveal that the proposed method improves enhancement performance by considering both the magnitude and phase of the complex spectrogram.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.1
• /
• pp.22-27
• /
• 2020

Recently, as the number of households raising pets increases due to the increase of single households, there is a need for a system for monitoring the status or behavior of pets. There are regional limitations in the monitoring of pets using domestic CCTVs, which requires a large number of CCTVs or restricts the behavior of pets. In this paper, we propose a mobile system for detecting and tracking cats using deep learning to solve the regional limitations of pet monitoring. We use YOLO (You Look Only Once), an object detection neural network model, to learn the characteristics of pets and apply them to Raspberry Pi to track objects detected in an image. We have designed a mobile monitoring system that connects Raspberry Pi and a laptop via wireless LAN and can check the movement and condition of cats in real time.

• Journal of the Korea Convergence Society
• /
• v.12 no.3
• /
• pp.1-7
• /
• 2021

Since Neural Machine Translation (NMT) uses only a limited number of words, there is a possibility that words that are not registered in the dictionary will be entered as input. The proposed method to alleviate this Out of Vocabulary (OOV) problem is Subword Tokenization, which is a methodology for constructing words by dividing sentences into subword units smaller than words. In this paper, we deal with general subword tokenization algorithms. Furthermore, in order to create a vocabulary that can handle the infinite conjugation of Korean adjectives and verbs, we propose a new methodology for subword tokenization training by separating the Jongsung(coda) from Korean syllables (consisting of Chosung-onset, Jungsung-neucleus and Jongsung-coda). As a result of the experiment, the methodology proposed in this paper outperforms the existing subword tokenization methodology.

• Smart Media Journal
• /
• v.10 no.1
• /
• pp.63-69
• /
• 2021

When a person sees a sentence and understands the sentence, the person understands the sentence by reminiscent of the main word in the sentence as an image. Text-to-image is what allows computers to do this associative process. The previous deep learning-based text-to-image model extracts text features using Convolutional Neural Network (CNN)-Long Short Term Memory (LSTM) and bi-directional LSTM, and generates an image by inputting it to the GAN. The previous text-to-image model uses basic embedding in text feature extraction, and it takes a long time to train because images are generated using several modules. Therefore, in this research, we propose a method of extracting features by using the attention mechanism, which has improved performance in the natural language processing field, for sentence embedding, and generating an image by inputting the extracted features into the GAN. As a result of the experiment, the inception score was higher than that of the model used in the previous study, and when judged with the naked eye, an image that expresses the features well in the input sentence was created. In addition, even when a long sentence is input, an image that expresses the sentence well was created.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.3
• /
• pp.389-395
• /
• 2021

In this paper, we propose a target image exchange model to improve performance of the object tracking algorithm based on a Siamese network. The object tracking algorithm based on the Siamese network tracks the object by finding the most similar part in the search image using only the target image specified in the first frame of the sequence. Since only the object of the first frame and the search image compare similarity, if tracking fails once, errors accumulate and drift in a part other than the tracked object occurs. Therefore, by designing a CNN(Convolutional Neural Network) based model, we check whether the tracking is progressing well, and the target image exchange timing is defined by using the score output from the Siamese network-based object tracking algorithm. The proposed model is evaluated the performance using the VOT-2018 dataset, and finally achieved an accuracy of 0.611 and a robustness of 22.816.