• Journal of the Korea Society of Computer and Information
• /
• v.24 no.9
• /
• pp.21-27
• /
• 2019

In this paper, we propose a deep auto-encoder-based pipe leak detection (PLD) technique from time-series acoustic data collected by microphone sensor nodes. The key idea of the proposed technique is to learn representative features of the leak-free state using leak-free time-series acoustic data and the deep auto-encoder. The proposed technique can be used to create a PLD model that detects leaks in the pipeline in an unsupervised learning manner. This means that we only use leak-free data without labeling while training the deep auto-encoder. In addition, when compared to the previous supervised learning-based PLD method that uses image features, this technique does not require complex preprocessing of time-series acoustic data owing to the unsupervised feature extraction scheme. The experimental results show that the proposed PLD method using the deep auto-encoder can provide reliable PLD accuracy even considering unsupervised learning-based feature extraction.

• Journal of Forest and Environmental Science
• /
• v.40 no.1
• /
• pp.15-23
• /
• 2024

This research aimed to assess the possibility of detecting forest degradation using time-series satellite imagery and three different deep learning-based change detection techniques. The dataset used for the deep learning models was composed of two sets, one based on surface reflectance (SR) spectral information from satellite imagery, combined with Texture Information (GLCM; Gray-Level Co-occurrence Matrix) and terrain information. The deep learning models employed for land cover change detection included image differencing using the Unet semantic segmentation model, multi-encoder Unet model, and multi-encoder Unet++ model. The study found that there was no significant difference in accuracy between the deep learning models for forest degradation detection. Both training and validation accuracies were approx-imately 89% and 92%, respectively. Among the three deep learning models, the multi-encoder Unet model showed the most efficient analysis time and comparable accuracy. Moreover, models that incorporated both texture and gradient information in addition to spectral information were found to have a higher classification accuracy compared to models that used only spectral information. Overall, the accuracy of forest degradation extraction was outstanding, achieving 98%.

• International journal of advanced smart convergence
• /
• v.12 no.4
• /
• pp.142-146
• /
• 2023

We present a method for generating 3D structures and rendering objects by combining VAE (Variational Autoencoder) and GAN (Generative Adversarial Network). This approach focuses on generating and rendering 3D models with improved quality using residual learning as the learning method for the encoder. We deep stack the encoder layers to accurately reflect the features of the image and apply residual blocks to solve the problems of deep layers to improve the encoder performance. This solves the problems of gradient vanishing and exploding, which are problems when constructing a deep neural network, and creates a 3D model of improved quality. To accurately extract image features, we construct deep layers of the encoder model and apply the residual function to learning to model with more detailed information. The generated model has more detailed voxels for more accurate representation, is rendered by adding materials and lighting, and is finally converted into a mesh model. 3D models have excellent visual quality and accuracy, making them useful in various fields such as virtual reality, game development, and metaverse.

• Journal of Information Processing Systems
• /
• v.20 no.1
• /
• pp.13-23
• /
• 2024

Music brings pleasure and relaxation to people. Therefore, it is necessary to classify musical genres based on scenes. Identifying favorite musical genres from massive music data is a time-consuming and laborious task. Recent studies have suggested that machine learning algorithms are effective in distinguishing between various musical genres. However, meeting the actual requirements in terms of accuracy or timeliness is challenging. In this study, a hybrid machine learning model that combines a deep residual auto-encoder (DRAE) and support vector machine (SVM) for musical genre recognition was proposed. Eight manually extracted features from the Mel-frequency cepstral coefficients (MFCC) were employed in the preprocessing stage as the hybrid music data source. During the training stage, DRAE was employed to extract feature maps, which were then used as input for the SVM classifier. The experimental results indicated that this method achieved a 91.54% F1-score and 91.58% top-1 accuracy, outperforming existing approaches. This novel approach leverages deep architecture and conventional machine learning algorithms and provides a new horizon for musical genre classification tasks.

• Journal of Korea Multimedia Society
• /
• v.21 no.8
• /
• pp.952-959
• /
• 2018

Clustering is one of the most fundamental algorithms in machine learning. The performance of clustering is affected by the distribution of data, and when there are more data or more dimensions, the performance is degraded. For this reason, we use a stacked auto encoder, one of the deep learning algorithms, to reduce the dimension of data which generate a feature vector that best represents the input data. We use k-means, which is a famous algorithm, as a clustering. Sine the feature vector which reduced dimensions are also multi dimensional, we use the Euclidean distance as well as the cosine similarity to increase the performance which calculating the similarity between the center of the cluster and the data as a vector. A deep clustering networks combining a stacked auto encoder and k-means re-trains the networks when the k-means result changes. When re-training the networks, the loss function of the stacked auto encoder and the loss function of the k-means are combined to improve the performance and the stability of the network. Experiments of benchmark image ad document dataset empirically validated the power of the proposed algorithm.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.2
• /
• pp.9-17
• /
• 2023

In this paper, we propose a one-class vibration anomaly detection system for bearing defect diagnosis. In order to reduce the economic and time loss caused by bearing failure, an accurate defect diagnosis system is essential, and deep learning-based defect diagnosis systems are widely studied to solve the problem. However, it is difficult to obtain abnormal data in the actual data collection environment for deep learning learning, which causes data bias. Therefore, a one-class classification method using only normal data is used. As a general method, the characteristics of vibration data are extracted by learning the compression and restoration process through AutoEncoder. Anomaly detection is performed by learning a one-class classifier with the extracted features. However, this method cannot efficiently extract the characteristics of the vibration data because it does not consider the frequency characteristics of the vibration data. To solve this problem, we propose an AutoEncoder model that considers the frequency characteristics of vibration data. As for classification performance, accuracy 0.910, precision 1.0, recall 0.820, and f1-score 0.901 were obtained. The network design considering the vibration characteristics confirmed better performance than existing methods.

• KIISE Transactions on Computing Practices
• /
• v.23 no.4
• /
• pp.268-273
• /
• 2017

Recently, deep learning encoder-based approach has been actively applied in the field of sentiment classification. However, Long Short-Term Memory network deep learning encoder, the commonly used architecture, lacks the quality of vector representation when the length of the documents is prolonged. In this study, for effective classification of the sentiment documents, we suggest the use of attention method-based deep learning encoder that generates document vector representation by weighted sum of the outputs of Long Short-Term Memory network based on importance. In addition, we propose methods to modify the attention method-based deep learning encoder to suit the sentiment classification field, which consist of a part that is to applied to window attention method and an attention weight adjustment part. In the window attention method part, the weights are obtained in the window units to effectively recognize feeling features that consist of more than one word. In the attention weight adjustment part, the learned weights are smoothened. Experimental results revealed that the performance of the proposed method outperformed Long Short-Term Memory network encoder, showing 89.67% in accuracy criteria.

• Journal of Korean Society of Water and Wastewater
• /
• v.38 no.2
• /
• pp.83-93
• /
• 2024

In this study, we present a sewer pipe inspection technique through a combination of active sonar technology and deep learning algorithms. It is difficult to inspect pipes containing water using conventional CCTV inspection methods, and there are various limitations, so a new approach is needed. In this paper, we introduce a inspection method using active sonar, and apply an auto encoder deep learning model to process sonar data to distinguish between normal and abnormal pipelines. This model underwent training on sonar data from a controlled environment under the assumption of normal pipeline conditions and utilized anomaly detection techniques to identify deviations from established standards. This approach presents a new perspective in pipeline inspection, promising to reduce the time and resources required for sewer system management and to enhance the reliability of pipeline inspections.

• Journal of Digital Convergence
• /
• v.19 no.6
• /
• pp.251-258
• /
• 2021

In this paper, we proposed an encoder-decoder model utilizing residual learning to improve the accuracy of the U-Net-based semantic segmentation method. U-Net is a deep learning-based semantic segmentation method and is mainly used in applications such as autonomous vehicles and medical image analysis. The conventional U-Net occurs loss in feature compression process due to the shallow structure of the encoder. The loss of features causes a lack of context information necessary for classifying objects and has a problem of reducing segmentation accuracy. To improve this, The proposed method efficiently extracted context information through an encoder using residual learning, which is effective in preventing feature loss and gradient vanishing problems in the conventional U-Net. Furthermore, we reduced down-sampling operations in the encoder to reduce the loss of spatial information included in the feature maps. The proposed method showed an improved segmentation result of about 12% compared to the conventional U-Net in the Cityscapes dataset experiment.

• The Journal of Korea Robotics Society
• /
• v.14 no.1
• /
• pp.8-13
• /
• 2019

Visual place recognition is widely researched area in robotics, as it is one of the elemental requirements for autonomous navigation, simultaneous localization and mapping for mobile robots. However, place recognition in changing environment is a challenging problem since a same place look different according to the time, weather, and seasons. This paper presents a feature extraction method using a deep convolutional auto-encoder to recognize places under severe appearance changes. Given database and query image sequences from different environments, the convolutional auto-encoder is trained to predict the images of the desired environment. The training process is performed by minimizing the loss function between the predicted image and the desired image. After finishing the training process, the encoding part of the structure transforms an input image to a low dimensional latent representation, and it can be used as a condition-invariant feature for recognizing places in changing environment. Experiments were conducted to prove the effective of the proposed method, and the results showed that our method outperformed than existing methods.