• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.3
• /
• pp.401-406
• /
• 2021

Recently, mean-teacher models based on convolutional recurrent neural networks are popularly used in audio event detection. The mean-teacher model is an architecture that consists of two parallel CRNNs and it is possible to train them effectively on the weakly-labelled and unlabeled audio data by using the consistency learning metric at the output of the two neural networks. In this study, we tried to improve the performance of the mean-teacher model by using additional derivative features of the log-mel spectrum. In the audio event detection experiments using the training and test data from the Task 4 of the DCASE 2018/2019 Challenges, we could obtain maximally a 8.1% relative decrease in the ER(Error Rate) in the mean-teacher model using proposed derivative features.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.2
• /
• pp.139-147
• /
• 2021

This paper introduces and implements a Sound Event Detection (SED) system based on weakly-supervised learning where only part of the data is labeled, and analyzes the effect of parameters. The SED system estimates the classes and onset/offset times of events in the acoustic signal. In order to train the model, all information on the event class and onset/offset times must be provided. Unfortunately, the onset/offset times are hard to be labeled exactly. Therefore, in the weakly-supervised task, the SED model is trained by "strongly labeled data" including the event class and activations, "weakly labeled data" including the event class, and "unlabeled data" without any label. Recently, the SED systems using the mean-teacher model are widely used for the task with several parameters. These parameters should be chosen carefully because they may affect the performance. In this paper, performance analysis was performed on parameters, such as the feature, moving average parameter, weight of the consistency cost function, ramp-up length, and maximum learning rate, using the data of DCASE 2020 Task 4. Effects and the optimal values of the parameters were discussed.

• Smart Media Journal
• /
• v.6 no.3
• /
• pp.41-48
• /
• 2017

Event detection using social media has been widespread since social network services have been an active communication channel for connecting with others, diffusing news message. Especially, the real-time characteristic of social media has created the opportunity for supporting for real-time applications/systems. Social network such as Twitter is the potential data source to explore useful information by mining messages posted by the user community. This paper proposed a novel system for temporal event detection by analyzing social data. As a result, this information can be used by first responders, decision makers, or news agents to gain insight of the situation. The proposed approach takes advantages of deep learning methods that play core techniques on the main tasks including informative data identifying from a noisy environment and temporal event detection. The former is the responsibility of Convolutional Neural Network model trained from labeled Twitter data. The latter is for event detection supported by Recurrent Neural Network module. We demonstrated our approach and experimental results on the case study of earthquake situations. Our system is more adaptive than other systems used traditional methods since deep learning enables to extract the features of data without spending lots of time constructing feature by hand. This benefit makes our approach adaptive to extend to a new context of practice. Moreover, the proposed system promised to respond to acceptable delay within several minutes that will helpful mean for supporting news channel agents or belief plan in case of disaster events.

• The Journal of Information Systems
• /
• v.27 no.4
• /
• pp.115-128
• /
• 2018

Purpose The purpose of this study is to predict future behaviors of business process. Specifically, this study tried to predict the last activities of process instances. It contributes to overcoming the limitations of existing approaches that they do not accurately reflect the actual behavior of business process and it requires a lot of effort and time every time they are applied to specific processes. Design/methodology/approach This study proposed a novel approach based using deep learning in the form of dynamic recurrent neural networks. To improve the accuracy of our prediction model based on the approach, we tried to adopt the latest techniques including new initialization functions(Xavier and He initializations). The proposed approach has been verified using real-life data of a domestic small and medium-sized business. Findings According to the experiment result, our approach achieves better prediction accuracy than the latest approach based on the static recurrent neural networks. It is also proved that much less effort and time are required to predict the behavior of business processes.

• Nuclear Engineering and Technology
• /
• v.52 no.9
• /
• pp.2009-2016
• /
• 2020

A nuclear power plant is a large complex system with tens of thousands of components. To ensure plant safety, the early and accurate diagnosis of abnormal situations is an important factor. To prevent misdiagnosis, operating procedures provide the anticipated symptoms of abnormal situations. While the more severe emergency situations total less than ten cases and can be diagnosed by dozens of key plant parameters, abnormal situations on the other hand include hundreds of cases and a multitude of parameters that should be considered for diagnosis. The tasks required of operators to select the appropriate operating procedure by monitoring large amounts of information within a limited amount of time can burden operators. This paper aims to develop a system that can, in a short time and with high accuracy, select the appropriate operating procedure and sub-procedure in an abnormal situation. Correspondingly, the proposed model has two levels of prediction to determine the procedure level and the detailed cause of an event. Simulations were conducted to evaluate the developed model, with results demonstrating high levels of performance. The model is expected to reduce the workload of operators in abnormal situations by providing the appropriate procedure to ultimately improve plant safety.

• Nuclear Engineering and Technology
• /
• v.54 no.10
• /
• pp.3620-3630
• /
• 2022

The diagnosis of abnormalities in a nuclear power plant is essential to maintain power plant safety. When an abnormal event occurs, the operator diagnoses the event and selects the appropriate abnormal operating procedures and sub-procedures to implement the necessary measures. To support this, abnormality diagnosis systems using data-driven methods such as artificial neural networks and convolutional neural networks have been developed. However, data-driven models cannot always guarantee an accurate diagnosis because they cannot simulate all possible abnormal events. Therefore, abnormality diagnosis systems should be able to detect their own potential misdiagnosis. This paper proposes a rulebased diagnostic validation algorithm using a previously developed two-stage diagnosis model in abnormal situations. We analyzed the diagnostic results of the sub-procedure stage when the first diagnostic results were inaccurate and derived a rule to filter the inconsistent sub-procedure diagnostic results, which may be inaccurate diagnoses. In a case study, two abnormality diagnosis models were built using gated recurrent units and long short-term memory cells, and consistency checks on the diagnostic results from both models were performed to detect any inconsistencies. Based on this, a re-diagnosis was performed to select the label of the second-best value in the first diagnosis, after which the diagnosis accuracy increased. That is, the model proposed in this study made it possible to detect diagnostic failures by the developed consistency check of the sub-procedure diagnostic results. The consistency check process has the advantage that the operator can review the results and increase the diagnosis success rate by performing additional re-diagnoses. The developed model is expected to have increased applicability as an operator support system in terms of selecting the appropriate AOPs and sub-procedures with re-diagnosis, thereby further increasing abnormal event diagnostic accuracy.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.195-199
• /
• 2021

본 논문에서는 재귀 신경망을 통해 동적 비전 센서 (DVS: Dynamic Vision Sensor)의 출력에서 엣지를 추정하는 방법을 제안한다. 동적 비전 센서는 기존의 일반적인 카메라들과 달리 급격한 움직임이나 밝기 변화에 강인하게 동작한다. 그러나 동적 비전 센서에서 획득한 출력은 각각이 독립적이기 때문에 화소들의 상관관계를 이용한 알고리즘을 사용함에 어려움이 따른다. 제안하는 방법은 센서에서 획득한 출력을 일정한 시간단위로 분할하고 2차원 평면에 투영함으로써 출력의 정보량 및 상관관계를 향상시키고, 이를 재귀 신경망에 통과시켜 엣지 정보를 추정한다. 이 방법은 센서의 출력에 의해 형성된 패턴을 학습하여 엣지를 잘 추출하였으며, 기존의 컴퓨터 비전 알고리즘의 적용 및 시각 관성 측위 등의 분야에서 활용될 수 있다.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.5
• /
• pp.479-487
• /
• 2021

In this paper, we propose an Sound Event Detection (SED) model using self-training based on a noisy student model. The proposed SED model consists of two stages. In the first stage, a mean-teacher model based on an Residual Convolutional Recurrent Neural Network (RCRNN) is constructed to provide target labels regarding weakly labeled or unlabeled data. In the second stage, a self-training-based noisy student model is constructed by applying different noise types. That is, feature noises, such as time-frequency shift, mixup, SpecAugment, and dropout-based model noise are used here. In addition, a semi-supervised loss function is applied to train the noisy student model, which acts as label noise injection. The performance of the proposed SED model is evaluated on the validation set of the Detection and Classification of Acoustic Scenes and Events (DCASE) 2020 Challenge Task 4. The experiments show that the single model and ensemble model of the proposed SED based on the noisy student model improve F1-score by 4.6 % and 3.4 % compared to the top-ranked model in DCASE 2020 challenge Task 4, respectively.

• Korean Journal of Social Welfare Studies
• /
• v.48 no.4
• /
• pp.221-249
• /
• 2017

This study aims to investigate determinants that affect job retention of injured workers depending upon types of return to work in order to suggest define the intervention priority for those who returned to original works and for those who did not. After constructing explaining variables based on literature reviews, determinants were verified analyzing 1,292 people using Panel Study of Worker's Compensation Insurance(PSWCI) data. The job retention period turned out to be 46.6 months for those who returned to original work and 34.2 month for those who returned to new works. Injured workers who return to new works tend to have more unemployment experiences. As a result of Cox proportional regression analysis, the longer it takes to return to work, the longer both groups tend to retain after the accident. Age, recuperation period, health status, psycho-social rehabilitation, education and occupational training also affect on job retention probability for those who return to new work. Based upon the analyzed result, setting up an adequate duration for return-to-work, intervention for injured workers who experienced vulnerable working condition before the accident and continuous case management after return-to-work are suggested.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.62 no.4
• /
• pp.1-12
• /
• 2020

The extreme 2017 spring drought affected a large portion of South Korea in the Southern Gyeonggi-do and Chungcheongnam-do districts. This drought event was one of the climatologically driest spring seasons over the 1961-2016 period of record. It was characterized by exceptionally low reservoir water levels, with the average water level being 36% lower over most of western South Korea. In this study, we consider drought response methods to alleviate the shortage of agricultural water in times of drought. It could be to store water from a stream into a reservoir. There is a cyclical method for reusing water supplied from a reservoir into streams through drainage. We intended to present a decision-making plan for water supply based on the calculation of the quantity of water supply and leakage. We compared the rainfall-runoff equation with the TANK model, which is a long-term run-off model. Estimations of reservoir inflow during non-irrigation seasons applied to the Madun, Daesa, and Pungjeon reservoirs. We applied the run-off flow to the last 30 years of rainfall data to estimate reservoir storage. We calculated the available water in the river during the non-irrigation season. The daily average inflow from 2003 to 2018 was calculated from October to April. Simulation results show that an average of 67,000 tons of water is obtained during the non-irrigation season. The report shows that about 53,000 tons of water are available except during the winter season from December to February. The Madun Reservoir began in early October with a 10 percent storage rate. In the starting ratio, a simulated rate of 4 K, 6 K, and 8 K tons is predicted to be 44%, 50%, and 60%. We can estimate the amount of water needed and the timing of water pump operations during the non-irrigation season that focuses on fresh water reservoirs and improve decision making for efficient water supplies.