• Journal of IKEEE
• /
• v.27 no.1
• /
• pp.59-64
• /
• 2023

Deep learning can solve various computer vision problems, but it requires a large dataset. Data augmentation technique based on image binarization for constructing large-scale datasets is proposed in this paper. By extracting features using image binarization and randomly placing the remaining pixels, new images are generated. The generated images showed similar quality to the original images and demonstrated excellent performance in deep learning models.

• Earthquakes and Structures
• /
• v.25 no.1
• /
• pp.15-26
• /
• 2023

It is of great importance to assess the residual displacement demand in the performance-based seismic design. In this paper, a hybrid deep learning model for predicting the residual displacement spectra under near-fault (NF) ground motions is proposed by combining the long short-term memory network (LSTM) and back-propagation (BP) network. The model is featured by its capacity of predicting the residual displacement spectrum under a given NF ground motion while considering the effects of structural parameters. To construct this model, 315 natural and artificial NF ground motions were employed to compute the residual displacement spectra through elastoplastic time history analysis considering different structural parameters. Based on the resulted dataset with a total of 9,450 samples, the proposed model was finally trained and tested. The results show that the proposed model has a satisfactory accuracy as well as a high efficiency in predicting residual displacement spectra under given NF ground motions while considering the impacts of structural parameters.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.153-155
• /
• 2022

Recently, we are using robots and drones to collect images. However, as the robot or drone is shaken by external influences, pre-processing technology to register images is required. Therefore, in this paper, we use autonomous robots, drones dataset and improve the quality of shaken image data through the Deblur deep learning model. We confirmed through the experimental results that the shaken images were registered and evaluated the model.

• Current Optics and Photonics
• /
• v.8 no.4
• /
• pp.391-398
• /
• 2024

An image reconstruction algorithm is vital for the image quality of a photonic integrated interferometric imaging (PIII) system. However, image reconstruction algorithms have limitations that always lead to degraded image reconstruction. In this paper, a novel image reconstruction algorithm based on deep learning is proposed. Firstly, the principle of optical signal transmission through the PIII system is investigated. A dataset suitable for image reconstruction of the PIII system is constructed. Key aspects such as model and loss functions are compared and constructed to solve the problem of image blurring and noise influence. By comparing it with other algorithms, the proposed algorithm is verified to have good reconstruction results not only qualitatively but also quantitatively.

• Journal of Korea Multimedia Society
• /
• v.23 no.9
• /
• pp.1129-1138
• /
• 2020

The performance of convolutional deep learning networks is generally determined according to parameters of target dataset, structure of network, convolution kernel, activation function, and optimization algorithm. In this paper, a genetic algorithm is used to select the appropriate deep learning model and parameters for Alzheimer's classification and to compare the learning results with preliminary experiment. We compare and analyze the Alzheimer's disease classification performance of VGG-16, GoogLeNet, and ResNet to select an effective network for detecting AD and MCI. The simulation results show that the network structure is ResNet, the activation function is ReLU, the optimization algorithm is Adam, and the convolution kernel has a 3-dilated convolution filter for the accuracy of dementia medical images.

• Journal of the Semiconductor & Display Technology
• /
• v.22 no.1
• /
• pp.113-117
• /
• 2023

As sputter equipment becomes more complex, it becomes increasingly difficult to understand the parameters that affect the thickness uniformity of thin metal film deposited by sputter. To address this issue, we verified a deep learning model that can predict complex relationships. Specifically, we trained the model to predict the height of 36 magnets based on the thickness of the material, using Support Vector Machine (SVM), Multilayer Perceptron (MLP), 1D-Convolutional Neural Network (1D-CNN), and 2D-Convolutional Neural Network (2D-CNN) algorithms. After evaluating each model, we found that the MLP model exhibited the best performance, especially when the dataset was constructed regardless of the thin film material. In conclusion, our study suggests that it is possible to predict the sputter equipment source using film thickness data through a deep learning model, which makes it easier to understand the relationship between film thickness and sputter equipment.

• The Korean Journal of Applied Statistics
• /
• v.37 no.2
• /
• pp.209-224
• /
• 2024

Speaker verification is becoming popular as a method of non-face-to-face identity authentication. It involves determining whether two voice data belong to the same speaker. In cases where the criminal's voice remains at the crime scene, it is vital to establish a speaker verification system that can accurately compare the two voice evidence. In this study, to achieve this, a new speaker verification system was built using a deep learning model for Korean language. High-dimensional voice data with a high variability like background noise made it necessary to use deep learning-based methods for speaker matching. To construct the matching algorithm, the ECAPA-TDNN model, known as the most famous deep learning system for speaker verification, was selected. A large dataset of the voice data, Voxceleb, collected from people of various nationalities without Korean. To study the appropriate form of datasets necessary for learning the Korean language, experiments were carried out to find out how Korean voice data affects the matching performance. The results showed that when comparing models learned only with Voxceleb and models learned with datasets combining Voxceleb and Korean datasets to maximize language and speaker diversity, the performance of learning data, including Korean, is improved for all test sets.

• International Journal of Computer Science & Network Security
• /
• v.21 no.5
• /
• pp.267-274
• /
• 2021

Medical care practices include gathering a wide range of student data that are with manic episodes and depression which would assist the specialist with diagnosing a health condition of the students correctly. In this way, the instructors of the specific students will also identify those students and take care of them well. The data which we collected from the students could be straightforward indications seen by them. The artificial intelligence has been utilized with Naive Baye's classification, Random forest classification algorithm, SVM algorithm to characterize the datasets which we gathered to check whether the student is influenced by Bipolar illness or not. Performance analysis of the disease data for the algorithms used is calculated and compared. Also, a sequential deep learning model is builded using Keras. The consequences of the simulations show the efficacy of the grouping techniques on a dataset, just as the nature and complexity of the dataset utilized.

• Applied Microscopy
• /
• v.50
• /
• pp.23.1-23.9
• /
• 2020

We propose an effective deep learning model to denoise scanning transmission electron microscopy (STEM) image series, named Noise2Atom, to map images from a source domain 𝓢 to a target domain 𝓒, where 𝓢 is for our noisy experimental dataset, and 𝓒 is for the desired clear atomic images. Noise2Atom uses two external networks to apply additional constraints from the domain knowledge. This model requires no signal prior, no noise model estimation, and no paired training images. The only assumption is that the inputs are acquired with identical experimental configurations. To evaluate the restoration performance of our model, as it is impossible to obtain ground truth for our experimental dataset, we propose consecutive structural similarity (CSS) for image quality assessment, based on the fact that the structures remain much the same as the previous frame(s) within small scan intervals. We demonstrate the superiority of our model by providing evaluation in terms of CSS and visual quality on different experimental datasets.

• ETRI Journal
• /
• v.43 no.4
• /
• pp.630-639
• /
• 2021

Autonomous driving requires a computerized perception of the environment for safety and machine-learning evaluation. Recognizing semantic information is difficult, as the objective is to instantly recognize and distinguish items in the environment. Training a model with real-time semantic capability and high reliability requires extensive and specialized datasets. However, generalized datasets are unavailable and are typically difficult to construct for specific tasks. Hence, a light detection and ranging semantic dataset suitable for semantic simultaneous localization and mapping and specialized for autonomous driving is proposed. This dataset is provided in a form that can be easily used by users familiar with existing two-dimensional image datasets, and it contains various weather and light conditions collected from a complex and diverse practical setting. An incremental and suggestive annotation routine is proposed to improve annotation efficiency. A model is trained to simultaneously predict segmentation labels and suggest class-representative frames. Experimental results demonstrate that the proposed algorithm yields a more efficient dataset than uniformly sampled datasets.