• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.242-252
• /
• 2021

Following the intuition that the local information in time instances is hardly incorporated into the posterior sequence in long short-term memory (LSTM), this paper proposes an attention augmented mechanism for fault diagnosis of the complex chemical process data. Unlike conventional fault diagnosis and classification methods, an attention mechanism layer architecture is introduced to detect and focus on local temporal information. The augmented deep network results preserve each local instance's importance and contribution and allow the interpretable feature representation and classification simultaneously. The comprehensive comparative analyses demonstrate that the developed model has a high-quality fault classification rate of 95.49%, on average. The results are comparable to those obtained using various other techniques for the Tennessee Eastman benchmark process.

• 한국인터넷방송통신학회논문지
• /
• 제23권6호
• /
• pp.89-95
• /
• 2023

인간 행동 인식 (Human action recognition, HAR) 기술은 스포츠 분석, 인간과 로봇 간의 상호작용, 대형 사이니지 콘텐츠 등의 애플리케이션에 활용되는 핵심 기술 중 하나이다. 본 논문에서는 몰입형 대형 사이니지 콘텐츠를 위한 STAGCN (Spatial temporal attention graph convolutional network) 기반 인간 행동 인식 시스템을 제안한다. STAGCN은 attention mechanism을 통해 스켈레톤 시퀀스의 시공간적 특징에 서로 다른 가중치를 부과하여, 동작 인식에 중요한 관절 및 시점을 고려할 수 있다. NTU RGB+D 데이터셋을 사용한 실험 결과, 제안된 시스템은 기존 딥러닝 모델들에 비해 높은 분류 정확도를 달성한 것을 확인했다.

• Journal of Information Processing Systems
• /
• 제15권5호
• /
• pp.1231-1242
• /
• 2019

This paper presents a two-dimensional attention-based long short-memory (2D-ALSTM) model for stock index prediction, incorporating input attention and temporal attention mechanisms for weighting of important stocks and important time steps, respectively. The proposed model is designed to overcome the long-term dependency, stock selection, and stock volatility delay problems that negatively affect existing models. The 2D-ALSTM model is validated in a comparative experiment involving the two attention-based models multi-input LSTM (MI-LSTM) and dual-stage attention-based recurrent neural network (DARNN), with real stock data being used for training and evaluation. The model achieves superior performance compared to MI-LSTM and DARNN for stock index prediction on a KOSPI100 dataset.

• Smart Structures and Systems
• /
• 제30권6호
• /
• pp.601-612
• /
• 2022

The prediction of structural mechanical behaviors is vital important to early perceive the abnormal conditions and avoid the occurrence of disasters. Especially for underground engineering, complex geological conditions make the structure more prone to disasters. Aiming at solving the problems existing in previous studies, such as incomplete consideration factors and can only predict the continuous performance, the deep attention fused temporal convolution network (DATCN) is proposed in this paper to predict the spatial mechanical behaviors of structure, which integrates both the temporal effect and spatial effect and realize the cross-time prediction. The temporal convolution network (TCN) and self-attention mechanism are employed to learn the temporal correlation of each monitoring point and the spatial correlation among different points, respectively. Then, the predicted result obtained from DATCN is compared with that obtained from some classical baselines, including SVR, LR, MLP, and RNNs. Also, the parameters involved in DATCN are discussed to optimize the prediction ability. The prediction result demonstrates that the proposed DATCN model outperforms the state-of-the-art baselines. The prediction accuracy of DATCN model after 24 hours reaches 90 percent. Also, the performance in last 14 hours plays a domain role to predict the short-term behaviors of the structure. As a study case, the proposed model is applied in an underwater shield tunnel to predict the stress variation of concrete segments in space.

• 한국멀티미디어학회논문지
• /
• 제14권3호
• /
• pp.378-391
• /
• 2011

본 논문에서는 입력장치로 들어오는 수많은 시각정보 중 현 시점에서 가장 유용하다고 생각되는 정보를 인간의 상향식 주의시각에 기반하여 선택하는 시각정보 선택기법에 대해 소개한다. 제안하는 시스템은 색상, 명도, 방위, 형태 등 저수준의 공간특징 외에 시간특징으로서 움직임 정보와 3차원 정보인 깊이 정보를 추가적으로 사용함으로써 기존방법에 비해 정보 선택의 정확도를 높혔다. 움직임 정보 추출 시 발생할 수 있는 노이즈를 제거하기 위해 인간의 움직임 인지에 대한 연구결과를 이용하는 새로운 접근법을 사용하였으며, 입력 영상 내 객체들이 부분적으로 겹쳐있다거나 동일한 현저도를 가지고 있을 때에도 현저한 영역을 제대로 선택해낼 수 있도록 깊이 정보를 사용하여 유의미한 영역을 선별하고 우선순위를 부여하였다. 실험결과를 통해 제안하는 방법이 기존의 방법에 비해 높은 정확도를 가짐을 확인할 수 있었다.

• 한국컴퓨터정보학회논문지
• /
• 제25권1호
• /
• pp.55-61
• /
• 2020

딥 러닝 기술의 발전과 컴퓨팅 파워 등의 개선으로 인해 비디오 기반 연구는 최근 많은 관심을 얻고 있다. 비디오 데이터가 이미지 데이터와 비교하여 가장 큰 차이는 비디오 데이터에는 많은 양의 시간적, 공간적 정보가 포함되어 있다는 점이다. 이처럼 비디오에 포함된 많은 양의 데이터로 인해 컴퓨터 비전 연구에 있어서 행동 인식은 중요한 연구 과제 중 하나이지만, 비디오와 같이 움직임이 있는 환경에서 인간의 행동 인식은 매우 복잡하고 도전적인 과제이다. 인간에 대한 여러 연구를 바탕으로 인공지능에서는 인간과 유사한 주의(attention)메커니즘이 효율적인 인식 모델이라는 것을 알게 되었다. 이 효율적인 모델은 이미지 정보와 복잡한 연속 비디오 정보를 처리하는 데 이상적이다. 본 논문에서는 이러한 연구배경을 기반으로, 비디오에서 인간의 행동을 효율적으로 인식하기 위해 먼저 인간의 행동에 주목한 후 비디오 행동 인식에 주의메커니즘을 도입하고자 한다. 논문의 주요내용은 두 가지 주의 메카니즘을 기반으로 컨볼루션 신경망을 이용한 새로운 3D 잔류 주의 네트워크를 제안함으로써 비디오에서 인간의 행동을 식별하고자 한다. 제안 모델의 평가 결과 최대 90.7%정도의 정확도를 보였다.

• Journal of Forest and Environmental Science
• /
• 제33권3호
• /
• pp.161-171
• /
• 2017

Global environmental changes have the capacity to make dramatic alterations to floral and faunal composition, and elucidation of the mechanism is important for predicting its outcomes. Studies on global climate change have traditionally focused on statistical summaries within relatively wide scales of spatial and temporal changes, and less attention has been paid to variability in microclimates across spatial and temporal scales. Microclimate is a suite of climatic conditions measured in local areas near the earth's surface. Environmental variables in microclimatic scale can be critical for the ecology of organisms inhabiting there. Here we examine the effect of spatial and temporal changes in microclimates on those of carabid beetle communities in Hyangnobong, Korea. We found that climatic variables and the patterns of annual changes in carabid beetle communities differed among sites even within the single mountain system. Our results indicate the importance of temporal survey of communities at local scales, which is expected to reveal an additional fraction of variation in communities and underlying processes that has been overlooked in studies of global community patterns and changes.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.1001-1004
• /
• 2021

Pain is an unpleasant experience for the patient. The recognition and assessment of pain help tailor the treatment to the patient, and they are also challenging in the medical. In this paper, we propose an approach for pain recognition through a deep neural network applied to pre-processed physiological. The proposed approach applies the idea of shortcut connections to concatenate the spatial information of a convolutional neural network and the temporal information of a recurrent neural network. In addition, our proposed approach applies the attention mechanism and achieves competitive performance on the BioVid Heat Pain dataset.

• 한국실내디자인학회논문집
• /
• 제25권6호
• /
• pp.116-126
• /
• 2016

In modern society which changes from quantity-seeking society to value-seeking one, people's various lifestyles have great effect on consumption patterns and work as an important factor in choosing hotels. The fact that design hotels, which provide unique experiences with differentiated and sensitive designs by reflecting various lifestyles, recently attract attention can be understood in the same context. As a matter of fact, design hotels recently serve as destinations as they become cultural and artistic icons which reflect customer lifestyles. Especially, the designs of lobby spaces in hotels play deciding role in customers' choices while representing the nature of hotels. In this respect, under the premise that the kinds of accumulated experiences are different depending on lifestyles and preferences for specific interior spaces are influenced by association mechanism formed by experiences, this study analyzed lobby spaces of design hotels which focus on specific lifestyles from the perspective of association mechanism based on experiences. As the method of analysis, this study classified the types of lifestyles and conducted case analysis to investigate what association mechanism works to enhance the preference of design hotels by types. Study classified lifestyles into experiential activity type, social meeting type, fashion-pursuing type and hideout-preferring type and analyzed cases of lobby designs in design hotels. The results of this case analysis are as follows; First, experiential activity type mainly utilized quasi-association and approach association through senses and social meeting type utilized quasi-association and memory association through emotions while fashion-pursuing type utilized quasi-association and presumption association through intuition and hideout-preferring type utilized quasi-association and approach association through thoughts. Second, it was found that most lobby designs are characterized by association mechanism in visual formative nature and that in temporal spatial nature working in complex way, and, through such process of association expansion, space stories are created. Stories of spaces created this way become unique identities of design hotels that provide new experiences for customers.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권6호
• /
• pp.1530-1544
• /
• 2023

Data monitoring is an important foundation of modern science. In most cases, the monitoring data is time-series data, which has high application value. The deep learning algorithm has a strong nonlinear fitting capability, which enables the recognition of time series by capturing anomalous information in time series. At present, the research of time series recognition based on deep learning is especially important for data monitoring. Deep learning algorithms require a large amount of data for training. However, abnormal sample is a small sample in time series, which means the number of abnormal time series can seriously affect the accuracy of recognition algorithm because of class imbalance. In order to increase the number of abnormal sample, a data augmentation method called GANBATS (GAN-based Bi-LSTM and Attention for Time Series) is proposed. In GANBATS, Bi-LSTM is introduced to extract the timing features and then transfer features to the generator network of GANBATS.GANBATS also modifies the discriminator network by adding an attention mechanism to achieve global attention for time series. At the end of discriminator, GANBATS is adding averagepooling layer, which merges temporal features to boost the operational efficiency. In this paper, four time series datasets and five data augmentation algorithms are used for comparison experiments. The generated data are measured by PRD(Percent Root Mean Square Difference) and DTW(Dynamic Time Warping). The experimental results show that GANBATS reduces up to 26.22 in PRD metric and 9.45 in DTW metric. In addition, this paper uses different algorithms to reconstruct the datasets and compare them by classification accuracy. The classification accuracy is improved by 6.44%-12.96% on four time series datasets.