• Journal of the Korean Society of Radiology
• /
• v.16 no.5
• /
• pp.575-583
• /
• 2022

Magnetic resonance diffusion tensor imaging (DTI) has revealed the disruption of brain white matter microstructure in normal aging and alcoholism undetectable with conventional structural MR imaging. we plan to analyze the FA measurements of the ROI of dangerous drinkers selected from Alcohol Use Disorders Identification Test (AUDIT) and Tract-Based Spatial Statics (TBSS) tool was used to extract FA values in the ROI from the image acquired through the pre-processing process. TBSS has a higher sensitivity of the FA value and MD value in the white matter than the brain gray matter, and has the advantage of quantitatively deriving the unlimited degree of brain nerve fibers, and more specialized in the brain white matter. We plan to analyze the fractional anisotropy (FA) measurement value for damage by selecting the center of the anatomical structure of the white matter region of the brain with high anisotropy among the brain neural networks that are particularly vulnerable to alcohol as the region of interest (ROI). In this study, we expected that alcohol causes damage to the brain white matter microstructure from FA value in various areas including both Choroid plexus. Especially, In the case of the moderate drunker, the mean value of FA in Lt, Rt. Choroid plexus was 0.2831 and 0.2872, whereas, in the case of the severe drunker, the mean value of FA was 0.1972 and 0.1936. We found that the higher the score on the AUDIT scale, the lower the FA value in ROI region of the brain white matter. Using the AUDIT scale, the guideline for the FA value of DTI can be presented, and it is possible to select a significant number of potentially severe drinkers. In other words, AUDIT was proved as useful tool in screening and discrimination of severe drunker through DTI.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.6
• /
• pp.1137-1144
• /
• 2022

In this study, an artificial intelligence(AI) was developed to help with facial expression practice in order to express emotions. The developed AI used multimodal inputs consisting of sentences and facial images for deep neural networks (DNNs). The DNNs calculated similarities between the emotions predicted by the sentences and the emotions predicted by facial images. The user practiced facial expressions based on the situation given by sentences, and the AI provided the user with numerical feedback based on the similarity between the emotion predicted by sentence and the emotion predicted by facial expression. ResNet34 structure was trained on FER2013 public data to predict emotions from facial images. To predict emotions in sentences, KoBERT model was trained in transfer learning manner using the conversational speech dataset for emotion classification opened to the public by AIHub. The DNN that predicts emotions from the facial images demonstrated 65% accuracy, which is comparable to human emotional classification ability. The DNN that predicts emotions from the sentences achieved 90% accuracy. The performance of the developed AI was evaluated through experiments with changing facial expressions in which an ordinary person was participated.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.49 no.2
• /
• pp.131-139
• /
• 2022

This study aimed to evaluate the effectiveness of deep convolutional neural networks (CNNs) for diagnosis of interproximal caries in pediatric intraoral radiographs. A total of 500 intraoral radiographic images of first and second primary molars were used for the study. A CNN model (Resnet 50) was applied for the detection of proximal caries. The diagnostic accuracy, sensitivity, specificity, receiver operating characteristic (ROC) curve, and area under ROC curve (AUC) were calculated on the test dataset. The diagnostic accuracy was 0.84, sensitivity was 0.74, and specificity was 0.94. The trained CNN algorithm achieved AUC of 0.86. The diagnostic CNN model for pediatric intraoral radiographs showed good performance with high accuracy. Deep learning can assist dentists in diagnosis of proximal caries lesions in pediatric intraoral radiographs.

• Journal of IKEEE
• /
• v.26 no.4
• /
• pp.633-638
• /
• 2022

Rencently, neuromorphic systems of spiking neural networks (SNNs) that imitate the human brain have attracted attention. Neuromorphic technology has the advantage of high speed and low power consumption in cognitive applications and processing. Resistive random-access memory (RRAM) for SNNs are the most efficient structure for parallel calculation and perform the gradual switching operation of spike-timing-dependent plasticity (STDP). RRAM as synaptic device operation has low-power processing and expresses various memory states. However, the integration of RRAM device causes high switching voltage and current, resulting in high power consumption. To reduce the operation voltage of the RRAM, it is important to develop new materials of the switching layer and metal electrode. This study suggested a optimized new structure that is the Metal/Al₂O₃/HfO_x/SWCNTs/N+silicon (MOCS) with single-walled carbon nanotubes (SWCNTs), which have excellent electrical and mechanical properties in order to lower the switching voltage. Therefore, we show an improvement in the gradual switching behavior and low-power I/V curve of SWCNTs-based memristors.

• Journal of Internet Computing and Services
• /
• v.23 no.2
• /
• pp.53-59
• /
• 2022

After the 3rd level autonomous driving vehicle, the 4th and 5th level of autonomous driving technology is trying to maintain the optimal condition of the passengers as well as the perfect driving of the vehicle. However current autonomous driving technology is too dependent on visual information such as LiDAR and front camera, so it is difficult to fully autonomously drive on roads other than designated roads. Therefore this paper proposes a Braking Strength Calculation System (BSCS), in which a vehicle classifies road conditions using data other than visual information and calculates optimal braking strength according to road conditions and driving conditions. The BSCS consists of RCDM (Road Condition Definition Module), which classifies road conditions based on KNN algorithm, and BSCM (Braking Strength Calculation Module), which calculates optimal braking strength while driving based on current driving conditions and road conditions. As a result of the experiment in this paper, it was possible to find the most suitable number of Ks for the KNN algorithm, and it was proved that the RCDM proposed in this paper is more accurate than the unsupervised K-means algorithm. By using not only visual information but also vibration data applied to the suspension, the BSCS of the paper can make the braking of autonomous vehicles smoother in various environments where visual information is limited.

• Journal of Intelligence and Information Systems
• /
• v.28 no.4
• /
• pp.347-364
• /
• 2022

Fraudulent companies or sellers strategically manipulate reviews to influence customers' purchase decisions; therefore, the reliability of reviews has become crucial for customer decision-making. Since customers increasingly rely on online reviews to search for more detailed information about products or services before purchasing, many researchers focus on detecting manipulated reviews. However, the main problem in detecting manipulated reviews is the difficulties with obtaining data with manipulated reviews to utilize machine learning techniques with sufficient data. Also, the number of manipulated reviews is insufficient compared with the number of non-manipulated reviews, so the class imbalance problem occurs. The class with fewer examples is under-represented and can hamper a model's accuracy, so machine learning methods suffer from the class imbalance problem and solving the class imbalance problem is important to build an accurate model for detecting manipulated reviews. Thus, we propose an OpenAI-based reviews generation model to solve the manipulated reviews imbalance problem, thereby enhancing the accuracy of manipulated reviews detection. In this research, we applied the novel autoregressive language model - GPT-3 to generate reviews based on manipulated reviews. Moreover, we found that applying GPT-3 model for oversampling manipulated reviews can recover a satisfactory portion of performance losses and shows better performance in classification (logit, decision tree, neural networks) than traditional oversampling models such as random oversampling and SMOTE.

• The Journal of the Convergence on Culture Technology
• /
• v.7 no.4
• /
• pp.675-681
• /
• 2021

With the recent growth of the AI speaker market, the demand for speech synthesis technology that enables natural conversation with users is increasing. Therefore, there is a need for a multi-speaker speech synthesis system that can generate voices of various tones. In order to synthesize natural speech, it is required to train with a large-capacity. high-quality speech DB. However, it is very difficult in terms of recording time and cost to collect a high-quality, large-capacity speech database uttered by many speakers. Therefore, it is necessary to train the speech synthesis system using the speech DB of a very large number of speakers with a small amount of training data for each speaker, and a technique for naturally expressing the tone and rhyme of multiple speakers is required. In this paper, we propose a technology for constructing a speaker encoder by applying the deep learning-based x-vector technique used in speaker recognition technology, and synthesizing a new speaker's tone with a small amount of data through the speaker encoder. In the multi-speaker speech synthesis system, the module for synthesizing mel-spectrogram from input text is composed of Tacotron2, and the vocoder generating synthesized speech consists of WaveNet with mixture of logistic distributions applied. The x-vector extracted from the trained speaker embedding neural networks is added to Tacotron2 as an input to express the desired speaker's tone.

• Journal of the Korea Computer Graphics Society
• /
• v.28 no.1
• /
• pp.1-9
• /
• 2022

Non-small cell lung cancer (NSCLC) accounts for a high proportion of 85% among all lung cancer and has a significantly higher mortality rate (22.7%) compared to other cancers. Therefore, it is very important to predict the prognosis after surgery in patients with non-small cell lung cancer. In this study, the types of preoperative chest CT image patches for non-small cell lung cancer patients with tumor as a region of interest are diversified into five types according to tumor-related information, and performance of single classifier model, ensemble classifier model with soft-voting method, and ensemble classifier model using 3 input channels for combination of three different patches using pre-trained ResNet and EfficientNet CNN networks are analyzed through misclassification cases and Grad-CAM visualization. As a result of the experiment, the ResNet152 single model and the EfficientNet-b7 single model trained on the peritumoral patch showed accuracy of 87.93% and 81.03%, respectively. In addition, ResNet152 ensemble model using the image, peritumoral, and shape-focused intratumoral patches which were placed in each input channels showed stable performance with an accuracy of 87.93%. Also, EfficientNet-b7 ensemble classifier model with soft-voting method using the image and peritumoral patches showed accuracy of 84.48%.

• Geophysics and Geophysical Exploration
• /
• v.25 no.2
• /
• pp.59-70
• /
• 2022

Recent studies demonstrate that machine learning has expanded in the field of seismic interpretation. Many convolutional neural networks have been developed for seismic sequence identification, which is important for seismic interpretation. However, expense and time limitations indicate that there is insufficient data available to provide a sufficient dataset to train supervised machine learning programs to identify seismic sequences. In this study, patch division and data augmentation are applied to mitigate this lack of data. Furthermore, to obtain spatial information that could be lost during patch division, an artificial channel is added to the original data to indicate depth. Seismic sequence identification is performed using a U-Net network and the Netherlands F3 block dataset from the dGB Open Seismic Repository, which offers datasets for machine learning, and the predicted results are evaluated. The results show that patch-based U-Net seismic sequence identification is improved by data augmentation and the addition of an artificial channel.

• Journal of Dental Rehabilitation and Applied Science
• /
• v.38 no.4
• /
• pp.196-203
• /
• 2022

Purpose: This study aimed to evaluate the accuracy and clinical usability of an identification model using deep learning for 79 dental implant types. Materials and Methods: A total of 45396 implant fixture images were collected through panoramic radiographs of patients who received implant treatment from 2001 to 2020 at 30 dental clinics. The collected implant images were 79 types from 18 manufacturers. EfficientNet and Meta Pseudo Labels algorithms were used. For EfficientNet, EfficientNet-B0 and EfficientNet-B4 were used as submodels. For Meta Pseudo Labels, two models were applied according to the widen factor. Top 1 accuracy was measured for EfficientNet and top 1 and top 5 accuracy for Meta Pseudo Labels were measured. Results: EfficientNet-B0 and EfficientNet-B4 showed top 1 accuracy of 89.4. Meta Pseudo Labels 1 showed top 1 accuracy of 87.96, and Meta pseudo labels 2 with increased widen factor showed 88.35. In Top5 Accuracy, the score of Meta Pseudo Labels 1 was 97.90, which was 0.11% higher than 97.79 of Meta Pseudo Labels 2. Conclusion: All four deep learning algorithms used for implant identification in this study showed close to 90% accuracy. In order to increase the clinical applicability of deep learning for implant identification, it will be necessary to collect a wider amount of data and develop a fine-tuned algorithm for implant identification.