• Journal of Korea Multimedia Society
• /
• v.21 no.8
• /
• pp.849-857
• /
• 2018

In this paper, we propose an automatic composition method using time series embedding of RNN Auto-Encoder. RNN Auto-Encoder can learn existing songs and can compose new songs from the trained RNN decoder. If one song is fully trained in the RNN Auto-Encoder, the song is embedded into the vector values of RNN nodes in the Auto-Encoder. If we train a lot of songs and apply a specific vector to the decoder of Auto-Encoder, then we can obtain a new song that combines the features of trained multiple songs according to the given vector. From extensive experiments we could find that our method worked well and generated various songs by selecting of the composition vectors.

• Journal of rehabilitation welfare engineering & assistive technology
• /
• v.11 no.2
• /
• pp.143-146
• /
• 2017

In this paper, a novel approach is proposed to improve the performance of speech/music classification using the recurrent neural network (RNN) in the enhanced voice services (EVS) of 3GPP for hearing aids. Feature vectors applied to the RNN are selected from the relevant parameters of the EVS for efficient speech/music classification. The performance of the proposed algorithm is evaluated under various conditions and large speech/music data. The proposed algorithm yields better results compared with the conventional scheme implemented in the EVS.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.456-457
• /
• 2021

Recently, in order to more accurately provide indoor location-based services, technologies using Wi-Fi fingerprints and deep learning are being studied. Among the deep learning models, an RNN model that can store information from the past can store continuous movements in indoor positioning, thereby reducing positioning errors. When using an RNN model for indoor positioning, the collected training data must be continuous sequential data. However, the Wi-Fi fingerprint data collected to determine specific location information cannot be used as training data for an RNN model because only RSSI for a specific location is recorded. This paper proposes a region clustering technique for sequential input data generation of RNN models based on Wi-Fi fingerprint data.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.34 no.4
• /
• pp.109-118
• /
• 2022

The data of the missing section among the vertex surface sea temperature observation data was imputed using the Bidirectional Recurrent Neural Network(BiRNN). Among artificial intelligence techniques, Recurrent Neural Networks (RNNs), which are commonly used for time series data, only estimate in the direction of time flow or in the reverse direction to the missing estimation position, so the estimation performance is poor in the long-term missing section. On the other hand, in this study, estimation performance can be improved even for long-term missing data by estimating in both directions before and after the missing section. Also, by using all available data around the observation point (sea surface temperature, temperature, wind field, atmospheric pressure, humidity), the imputation performance was further improved by estimating the imputation data from these correlations together. For performance verification, a statistical model, Multivariate Imputation by Chained Equations (MICE), a machine learning-based Random Forest model, and an RNN model using Long Short-Term Memory (LSTM) were compared. For imputation of long-term missing for 7 days, the average accuracy of the BiRNN/statistical models is 70.8%/61.2%, respectively, and the average error is 0.28 degrees/0.44 degrees, respectively, so the BiRNN model performs better than other models. By applying a temporal decay factor representing the missing pattern, it is judged that the BiRNN technique has better imputation performance than the existing method as the missing section becomes longer.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.1
• /
• pp.774-779
• /
• 2020

The seismic response reduction capacity of a smart mid-story isolation system was investigated using the RNN model in this study. For this purpose, an RNN model was developed to make a dynamic response prediction of building structures subjected to seismic loads. An existing tall building with a mid-story isolation system was selected as an example structure for realistic research. A smart mid-story isolation system was comprised of an MR damper instead of existing lead dampers. The RNN model predicted the seismic responses accurately compared to those of the FEM model. The simulation time of the RNN model can be reduced significantly compared to the FEM model. After the numerical simulations, the smart mid-story isolation system could effectively reduce the seismic responses of the existing building compared to the conventional mid-story isolation system.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2018.07a
• /
• pp.360-363
• /
• 2018

본 논문에서는 계층적 RNN의 성능 향상을 위하여 강화학습을 통한 계층적 RNN 내 파라미터를 효율적으로 찾는 방법을 제안한다. 계층적 RNN 내 임의의 파라미터에서 학습을 진행하고 얻는 분류 정확도를 보상으로 하여 간소화된 강화학습 네트워크에서 보상을 최대화하도록 강화학습 내부 파라미터를 수정한다. 기존의 강화학습을 통한 내부 구조를 찾는 네트워크는 많은 자원과 시간을 소모하므로 이를 해결하기 위해 간소화된 강화학습 구조를 적용하였고 이를 통해 적은 컴퓨터 자원에서 학습속도를 증가시킬 수 있었다. 간소화된 강화학습을 통해 계층적 RNN의 파라미터를 수정하고 이를 행동 인식 데이터 세트에 적용한 결과 기존 알고리즘 대비 높은 성능을 얻을 수 있었다.

• Journal of KIISE
• /
• v.44 no.6
• /
• pp.607-612
• /
• 2017

This paper presents an attempt to apply Deep Learning technology to solve the problem of forecasting floods in urban areas. We employ Recurrent Neural Networks (RNNs), which are suitable for analyzing time series data, to learn observed data of river water and to predict the water level. To test the model, we use water observation data of a station in the Trinity river, Texas, the U.S., with data from 2013 to 2015 for training and data in 2016 for testing. Input of the neural networks is a 16-record-length sequence of 15-minute-interval time-series data, and output is the predicted value of the water level at the next 30 minutes and 60 minutes. In the experiment, we compare three Deep Learning models including standard RNN, RNN trained with Back Propagation Through Time (RNN-BPTT), and Long Short-Term Memory (LSTM). The prediction quality of LSTM can obtain Nash Efficiency exceeding 0.98, while the standard RNN and RNN-BPTT also provide very high accuracy.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.243-246
• /
• 2018

최근 사용자 발화를 이해하고 그에 맞는 피드백을 생성할 수 있는 대화 시스템의 중요성이 증가하고 있다. 따라서 사용자 의도를 파악하기 위한 화행 분석은 대화 시스템의 필수적인 요소이다. 최근 많이 연구되는 심층 학습 기법은 모델이 데이터로부터 자질들을 스스로 추출한다는 장점이 있다. 발화 자체의 연속성과 화자간 상호 작용을 포착하기 위하여 CNN에 RNN을 결합한 CNN-RNN을 제안한다. 본 논문에서 제안한 계층 구조 어텐션 매커니즘 기반 CNN-RNN을 효과적으로 적용한 결과 워드 임베딩을 추가한 조건에서 가장 높은 성능인 91.72% 정확도를 얻었다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.5
• /
• pp.533-539
• /
• 2019

This study proposes a malware classification model that can handle arbitrary length input data using the Microsoft Malware Classification Challenge dataset. We are based on imaging existing data from malware. The proposed model generates a lot of images when malware data is large, and generates a small image of small data. The generated image is learned as time series data by Dynamic RNN. The output value of the RNN is classified into malware by using only the highest weighted output by applying the Attention technique, and learning the RNN output value by Residual CNN again. Experiments on the proposed model showed a Micro-average F1 score of 92% in the validation data set. Experimental results show that the performance of a model capable of learning and classifying arbitrary length data can be verified without special feature extraction and dimension reduction.

• Journal of Internet Computing and Services
• /
• v.22 no.3
• /
• pp.67-73
• /
• 2021

In this paper, in order to prevent the fall of the drone, a study was conducted to collect vibration data from the motor connected to the propeller of the drone, and to predict the abnormal vibration of the drone using recurrent neural network (RNN) and long short term memory (LSTM). In order to collect the vibration data of the drone, a vibration sensor is attached to the motor connected to the propeller of the drone to collect vibration data on normal, bar damage, rotor damage, and shaft deflection, and abnormal vibration data are collected through LSTM and RNN. The root mean square error (RMSE) value of the vibration prediction result were compared and analyzed. As a result of the comparative simulation, it was confirmed that both the predicted result through RNN and LSTM predicted the abnormal vibration pattern very accurately. However, the vibration predicted by the LSTM was found to be 15.4% lower on average than the vibration predicted by the RNN.