• The Journal of the Acoustical Society of Korea
• /
• v.42 no.4
• /
• pp.345-356
• /
• 2023

Target signal of passive sonar shows narrow band harmonic characteristic with a variation in intensity within a few seconds and long term frequency variation due to the Lloyd's mirror effect. We propose a signal classification algorithm based on Gated Recurrent Unit (GRU) that learns local and global time series features. The algorithm proposed implements a multi layer network using GRU and extracts local and global time series features via dilated connections. We learns attention mechanism to weight time series features and classify passive sonar signals. In experiments using public underwater acoustic data, the proposed network showed superior classification accuracy of 96.50 %. This result is 4.17 % higher classification accuracy compared to existing skip connected GRU network.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.6
• /
• pp.536-543
• /
• 2023

Speaker diarization, which labels for "who spoken when?" in speech with multiple speakers, has been studied on a deep neural network-based end-to-end method for labeling on speech overlap and optimization of speaker diarization models. Most deep neural network-based end-to-end speaker diarization systems perform multi-label classification problem that predicts the labels of all speakers spoken in each frame of speech. However, the performance of the multi-label-based model varies greatly depending on what the threshold is set to. In this paper, it is studied a speaker diarization system using single-label classification so that speaker diarization can be performed without thresholds. The proposed model estimate labels from the output of the model by converting speaker labels into a single label. To consider speaker label permutations in the training, the proposed model is used a combination of Permutation Invariant Training (PIT) loss and cross-entropy loss. In addition, how to add the residual connection structures to model is studied for effective learning of speaker diarization models with deep structures. The experiment used the Librispech database to generate and use simulated noise data for two speakers. When compared with the proposed method and baseline model using the Diarization Error Rate (DER) performance the proposed method can be labeling without threshold, and it has improved performance by about 20.7 %.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.9
• /
• pp.395-403
• /
• 2024

Recently, with the advancement of technology, the automotive industry has seen an increase in network connectivity. CAN (Controller Area Network) bus technology enables fast and efficient data communication between various electronic devices and systems within a vehicle, providing a platform that integrates and manages a wide range of functions, from core systems to auxiliary features. However, this increased connectivity raises concerns about network security, as external attackers could potentially gain access to the automotive network, taking control of the vehicle or stealing personal information. This paper analyzed abnormal messages occurring in CAN and confirmed that message occurrence periodicity, frequency, and data changes are important factors in the detection of abnormal messages. Through DBC decoding, the specific meanings of CAN messages were interpreted. Based on this, a model for classifying abnormalities was proposed using the GRU model to analyze the periodicity and trend of message occurrences by measuring the difference (residual) between the predicted and actual messages occurring within a certain period as an abnormality metric. Additionally, for multi-class classification of attack techniques on abnormal messages, a Random Forest model was introduced as a multi-classifier using message occurrence frequency, periodicity, and residuals, achieving improved performance. This model achieved a high accuracy of over 99% in detecting abnormal messages and demonstrated superior performance compared to other existing models.

• Journal of IKEEE
• /
• v.24 no.4
• /
• pp.1122-1128
• /
• 2020

Bandwidth Extension refers to restoring and expanding a narrow band signal(NB) that is damaged or damaged in the encoding and decoding process due to the lack of channel capacity or the characteristics of the codec installed in the mobile communication device. It means converting to a wideband signal(WB). Bandwidth extension research mainly focuses on voice signals and converts high bands into frequency domains, such as SBR (Spectral Band Replication) and IGF (Intelligent Gap Filling), and restores disappeared or damaged high bands based on complex feature extraction processes. In this paper, we propose a model that outputs an bandwidth extended signal based on an autoencoder among deep learning models, using the residual connection of one-dimensional convolutional neural networks (CNN), the bandwidth is extended by inputting a time domain signal of a certain length without complicated pre-processing. In addition, it was confirmed that the damaged high band can be restored even by training on a dataset containing various types of sound sources including music that is not limited to the speech.

• Journal of Practical Engineering Education
• /
• v.16 no.5_spc
• /
• pp.611-618
• /
• 2024

The rapid advancement of deep learning has significantly enhanced the performance of single image super-resolution (SR). However, most existing deep learning-based image SR networks only facilitate information flow in the forward direction, which limits their performance. In this study, we investigate a feedback network for precise image SR. This feedback network effectively enhances lower-level feature representation by rerouting multiple higher-level features. We sequentially construct several Residual Density Modules and deploy them repeatedly over time. Multiple feedback connections between two adjacent time steps leverage high-level features captured within a large receptive field to refine low-level features lacking sufficient contextual information. A carefully designed feedback module efficiently selects and enhances valuable information from the rerouted high-level features, thereby improving low-level features with enriched high-level information. Extensive experiments demonstrate that the proposed method outperforms existing approaches in both objective and subjective evaluations.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.5
• /
• pp.185-191
• /
• 2022

In this study, we propose a novel super-resolution neural network based on U-Net, residual neural network, and sub-pixel convolution. To prevent the loss of detailed information due to the max pooling of U-Net, we propose down-sampling and connection using sub-pixel convolution. This uses all pixels in the filter, unlike the max pooling that creates a new feature map with only the max value in the filter. As a 2×2 size filter passes, it creates a feature map consisting only of pixels in the upper left, upper right, lower left, and lower right. This makes it half the size and quadruple the number of feature maps. And we propose two methods to reduce the computation. The first uses sub-pixel convolution, which has no computation, and has better performance, instead of up-convolution. The second uses a layer that adds two feature maps instead of the connection layer of the U-Net. Experiments with a banchmark dataset show better PSNR values on all scale and benchmark datasets except for set5 data on scale 2, and well represent local area patterns.

• Journal of Korean Society for Geospatial Information Science
• /
• v.22 no.3
• /
• pp.47-56
• /
• 2014

Recently, the velocity of sea-level rising has increased due to the global warming and the natural disasters have been occurred many times. Therefore, there are various demands for the integration of vertical reference datums for the ocean and land areas in order to develop a coastal area and prevent a natural disaster. Currently, the vertical datum for the ocean area refers to Local Mean Sea Level(LMSL) and the vertical datum for the land area is based on Incheon Mean Sea Level(IMSL). This study uses 31 points of Tidal Gauge Bench Mark (TGBM) in order to compares and analyzes the geometric heights referring LMSL, IMSL, and the nationally determined geoid surface. 11 points of comparable data are biased more than 10 cm when the geometric heights are compared. It seems to be caused by the inflow of river, the relocation of Tidal Gauge Station, and the topographic change by harbor construction. Also, this study analyze the inclination of sea surface which is the difference between IMSL and LMSL, and it shows the inclination of sea surface increases from the western to southern, and eastern seas. In this study, it is shown that TGBM can be used to integrate vertical datums for the ocean and land areas. In order to integrate the vertical datums, there need more surveying data connecting the ocean to the land area, also cooperation between Korea Hydrographic and Oceanographic Administration and National Geographic Information Institute. It is expected that the integrated vertical datum can be applied to the development of coastal area and the preventative of natural disaster.

• Journal of Korea Water Resources Association
• /
• v.42 no.6
• /
• pp.445-455
• /
• 2009

In this paper proportional hazards models for the first through seventh break of 150 mm cast iron pipes in a case study area are established. During the modeling process the assumption of the proportional hazards for covariates on the hazards is examined to include the time-dependent covariate terms in the models. As a result, the pipe material/joint type and the number of customers are modeled as time-dependent for the first failure, and for the second failure only the number of customers is modeled as time-dependent. From the analysis on the baseline hazard functions the failure hazards are found to be generally increasing for the first and second failure, while the hazards of the third break and beyond showed a form of a bath-tub. Furthermore, the changes in the baseline hazard rates according to the time and number of break reflect that the general condition of the pipes is deteriorating. The factors causing pipe break and their effects are analyzed based on the estimated regression coefficients and their hazard ratios, and the constructed models are verified using the deviance residuals of the models.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.441-446
• /
• 2020

In this paper, we propose the α-Feature Map Scaling (α-FMS) method which extends the FMS method that was designed to enhance the discriminative power of feature maps of deep neural networks in Speaker Verification (SV) systems. The FMS derives a scale vector from a feature map and then adds or multiplies them to the features, or sequentially apply both operations. However, the FMS method not only uses an identical scale vector for both addition and multiplication, but also has a limitation that it can only add a value between zero and one in case of addition. In this study, to overcome these limitations, we propose α-FMS to add a trainable parameter α to the feature map element-wise, and then multiply a scale vector. We compare the performance of the two methods: the one where α is a scalar, and the other where it is a vector. Both α-FMS methods are applied after each residual block of the deep neural network. The proposed system using the α-FMS methods are trained using the RawNet2 and tested using the VoxCeleb1 evaluation set. The result demonstrates an equal error rate of 2.47 % and 2.31 % for the two α-FMS methods respectively.

• Journal of Broadcast Engineering
• /
• v.24 no.1
• /
• pp.132-141
• /
• 2019

The necessity of transmitting video data over a narrow-bandwidth exists steadily despite that video service over broadband is common. In this paper, we propose a scalable video coding framework for low-resolution video transmission over a very narrow-bandwidth network by super-resolution of decoded frames of a base layer using a convolutional neural network based super resolution technique to improve the coding efficiency by using it as a prediction for the enhancement layer. In contrast to the conventional scalable high efficiency video coding (SHVC) standard, in which upscaling is performed with a fixed filter, we propose a scalable video coding framework that replaces the existing fixed up-scaling filter by using the trained convolutional neural network for super-resolution. For this, we proposed a neural network structure with skip connection and residual learning technique and trained it according to the application scenario of the video coding framework. For the application scenario where a video whose resolution is $352{\times}288$ and frame rate is 8fps is encoded at 110kbps, the quality of the proposed scalable video coding framework is higher than that of the SHVC framework.