• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.4
• /
• pp.161-166
• /
• 2018

In this paper, we propose a method for the random selection of pooling operations for the regularization and reduction of cross validation in convolutional neural networks. The pooling operation in convolutional neural networks is used to reduce the size of the feature map and for its shift invariant properties. In the existing pooling method, one pooling operation is applied in each pooling layer. Because this method fixes the convolution network, the network suffers from overfitting, which means that it excessively fits the models to the training samples. In addition, to find the best combination of pooling operations to maximize the performance, cross validation must be performed. To solve these problems, we introduce the probability concept into the pooling layers. The proposed method does not select one pooling operation in each pooling layer. Instead, we randomly select one pooling operation among multiple pooling operations in each pooling region during training, and for testing purposes, we use probabilistic weighting to produce the expected output. The proposed method can be seen as a technique in which many networks are approximately averaged using a different pooling operation in each pooling region. Therefore, this method avoids the overfitting problem, as well as reducing the amount of cross validation. The experimental results show that the proposed method can achieve better generalization performance and reduce the need for cross validation.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.4
• /
• pp.129-136
• /
• 2020

PM2.5 which is a very tiny air particulate matter even smaller than PM10 has been issued in the environmental problem. Since PM2.5 can cause eye diseases or respiratory problems and infiltrate even deep blood vessels in the brain, it is important to predict PM2.5. However, it is difficult to predict PM2.5 because there is no clear explanation yet regarding the creation and the movement of PM2.5. Thus, prediction methods which not only predict PM2.5 accurately but also have the interpretability of the result are needed. To predict hourly PM2.5 of Seoul city, we propose a method using random forest with the adjusted bootstrap number from the time series ground data preprocessed on different sources. With this method, the prediction model can be trained uniformly on hourly information and the result has the interpretability. To evaluate the prediction performance, we conducted comparative experiments. As a result, the performance of the proposed method was superior against other models in all labels. Also, the proposed method showed the importance of the variables regarding the creation of PM2.5 and the effect of China.

• Journal of the Korea Convergence Society
• /
• v.12 no.7
• /
• pp.193-205
• /
• 2021

The AI speaker market is growing steadily. However, the satisfaction of actual users is only 42%. Therefore, in this paper, we collected reviews on Amazon Echo Dot 3rd and 4th generation models to analyze what hinders the user experience through the topic changes and emotional changes of each generation of AI speakers. By using topic modeling analysis techniques, we found changes in topics and topics that make up reviews for each generation, and examined how user sentiment on topics changed according to generation through deep learning-based sentiment analysis. As a result of topic modeling, five topics were derived for each generation. In the case of the 3rd generation, the topic representing general features of the speaker acted as a positive factor for the product, while user convenience features acted as negative factor. Conversely, in the 4^th generation, general features were negatively, and convenience features were positively derived. This analysis is significant in that it can present analysis results that take into account not only lexical features but also contextual features of the entire sentence in terms of methodology.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.3
• /
• pp.183-191
• /
• 2021

Non-intrusive load monitoring is a technology that can be used for predicting and classifying the type of appliances through real-time monitoring of user power consumption, and it has recently got interested as a means of energy-saving. In this paper, we propose a system for classifying appliances from user consumption data by combining GAF(Gramian angular field) technique that can be used for converting one-dimensional data to the two-dimensional matrix with convolutional neural networks. We use REDD(residential energy disaggregation dataset) that is the public appliances power data and confirm the classification accuracy of the GASF(Gramian angular summation field) and GADF(Gramian angular difference field). Simulation results show that both models showed 94% accuracy on appliances with binary-state(on/off) and that GASF showed 93.5% accuracy that is 3% higher than GADF on appliances with multi-state. In later studies, we plan to increase the dataset and optimize the model to improve accuracy and speed.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_1
• /
• pp.1141-1180
• /
• 2022

The application of satellite imageries has increased in the field of hydrology and water resources in recent years. However, challenges have been encountered on obtaining accurate evapotranspiration and soil moisture. Therefore, present researches have emphasized the necessity to obtain estimations of satellite-based evapotranspiration and soil moisture with related development researches. In this study, we presented the research status in Korea by investigating the current trends and methodologies for evapotranspiration and soil moisture. As a result of examining the detailed methodologies, we have ascertained that, in general, evapotranspiration is estimated using Energy balance models, such as Surface Energy Balance Algorithm for Land (SEBAL) and Mapping Evapotranspiration with Internalized Calibration (METRIC). In addition, Penman-Monteith and Priestley-Taylor equations are also used to estimate evapotranspiration. In the case of soil moisture, in general, active (AMSR-E, AMSR2, MIRAS, and SMAP) and passive (ASCAT and SAR)sensors are used for estimation. In terms of statistics, deep learning, as well as linear regression equations and artificial neural networks, are used for estimating these parameters. There were a number of research cases in which various indices were calculated using satellite-based data and applied to the characterization of drought. In some cases, hydrological cycle factors of evapotranspiration and soil moisture were calculated based on the Land Surface Model (LSM). Through this process, by comparing, reviewing, and presenting major detailed methodologies, we intend to use these references in related research, and lay the foundation for the advancement of researches on the calculation of satellite-based hydrological cycle data in the future.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.14 no.2
• /
• pp.128-133
• /
• 2021

Currently, autonomous vehicle markets are commercializing a third-level autonomous vehicle, but there is a possibility that an accident may occur even during fully autonomous driving due to stability issues. In fact, autonomous vehicles have recorded 81 accidents. This is because, unlike level 3, autonomous vehicles after level 4 have to judge and respond to emergency situations by themselves. Therefore, this paper proposes a vehicle crisis detection system(VCDS) that collects and stores information outside the vehicle through CNN, and uses the stored information and vehicle sensor data to output the crisis situation of the vehicle as a number between 0 and 1. The VCDS consists of two modules. The vehicle external situation collection module collects surrounding vehicle and pedestrian data using a CNN-based neural network model. The vehicle crisis situation determination module detects a crisis situation in the vehicle by using the output of the vehicle external situation collection module and the vehicle internal sensor data. As a result of the experiment, the average operation time of VESCM was 55ms, R-CNN was 74ms, and CNN was 101ms. In particular, R-CNN shows similar computation time to VESCM when the number of pedestrians is small, but it takes more computation time than VESCM as the number of pedestrians increases. On average, VESCM had 25.68% faster computation time than R-CNN and 45.54% faster than CNN, and the accuracy of all three models did not decrease below 80% and showed high accuracy.

• Journal of Korea Water Resources Association
• /
• v.55 no.12
• /
• pp.1115-1124
• /
• 2022

The impact of climate change on water resources was evaluated for Chungju Dam and Soyang-gang Dam basins by constructing an integrated modeling framework consisting of a dam inflow prediction model based on the Variable Infiltration Capacity (VIC) model, a distributed hydrologic model, and an LSTM based dam outflow prediction model. Considering the uncertainty of future climate data, four models of CMIP6 GCM were used as input data of VIC model for future period (2021-2100). As a result of applying future climate data, the average inflow for period increased as the future progressed, and the inflow in the far future (2070-2100) increased by up to 22% compared to that of the observation period (1986-2020). The minimum value of dam discharge lasting 4~50 days was significantly lower than the observed value. This indicates that droughts may occur over a longer period than observed in the past, meaning that citizens of Seoul metropolitan areas may experience severe water shortages due to future droughts. In addition, compared to the near and middle futures, the change in water storage has occurred rapidly in the far future, suggesting that the difficulties of water resource management may increase.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.359-366
• /
• 2022

In this paper, we present the Korean menu-ordering Sentence Text-to-Speech (TTS) system using conformer-based FastSpeech2. Conformer is the convolution-augmented transformer, which was originally proposed in Speech Recognition. Combining two different structures, the Conformer extracts better local and global features. It comprises two half Feed Forward module at the front and the end, sandwiching the Multi-Head Self-Attention module and Convolution module. We introduce the Conformer in Korean TTS, as we know it works well in Korean Speech Recognition. For comparison between transformer-based TTS model and Conformer-based one, we train FastSpeech2 and Conformer-based FastSpeech2. We collected a phoneme-balanced data set and used this for training our models. This corpus comprises not only general conversation, but also menu-ordering conversation consisting mainly of loanwords. This data set is the solution to the current Korean TTS model's degradation in loanwords. As a result of generating a synthesized sound using ParallelWave Gan, the Conformer-based FastSpeech2 achieved superior performance of MOS 4.04. We confirm that the model performance improved when the same structure was changed from transformer to Conformer in the Korean TTS.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.5
• /
• pp.274-286
• /
• 2022

Multi-object tracking has been studied for a long time under computer vision and plays a critical role in applications such as autonomous driving and driving assistance. Multi-object tracking techniques generally consist of a detector that detects objects and a tracker that tracks the detected objects. Various publicly available datasets allow us to train a detector model without much effort. However, there are relatively few publicly available datasets for training a tracker model, and configuring own tracker datasets takes a long time compared to configuring detector datasets. Hence, the detector is often developed separately with a tracker module. However, the separated tracker should be adjusted whenever the former detector model is changed. This study proposes a system that can train a model that performs detection and tracking simultaneously using only the detector training datasets. In particular, a Siam network with augmentation is used to compose the detector and tracker. Experiments are conducted on public datasets to verify that the proposed algorithm can formulate a real-time multi-object tracker comparable to the state-of-the-art tracker models.

• The Korean Journal of Applied Statistics
• /
• v.36 no.2
• /
• pp.101-114
• /
• 2023

The data augmentation technique is effectively used to solve the problem of overfitting the model by allowing the training dataset to be viewed from various perspectives. In addition to image augmentation techniques such as rotation, cropping, horizontal flip, and vertical flip, occlusion-based data augmentation methods such as Cutmix and Cutout have been proposed. For models based on speech data, it is possible to use an occlusion-based data-based augmentation technique after converting a 1D speech signal into a 2D spectrogram. In particular, SpecAugment is an occlusion-based augmentation technique for speech spectrograms. In this study, we intend to compare and study data augmentation techniques that can be used in the problem of false-voice detection. Using data from the ASVspoof2017 and ASVspoof2019 competitions held to detect fake audio, a dataset applied with Cutout, Cutmix, and SpecAugment, an occlusion-based data augmentation method, was trained through an LCNN model. All three augmentation techniques, Cutout, Cutmix, and SpecAugment, generally improved the performance of the model. In ASVspoof2017, Cutmix, in ASVspoof2019 LA, Mixup, and in ASVspoof2019 PA, SpecAugment showed the best performance. In addition, increasing the number of masks for SpecAugment helps to improve performance. In conclusion, it is understood that the appropriate augmentation technique differs depending on the situation and data.