• Therapeutic Science for Rehabilitation
• /
• v.6 no.2
• /
• pp.71-83
• /
• 2017

Objective: This study was to develop and verify the effects of the exercise-cognitive combined dual-task training program on cognitive function and depression of the elderly with mild cognitive impairment(MCI). Methods: The subjects were randomly assigned to the exercise-cognitive combined dual-task training group(n=32) or single-task training group(n=31). To identify the effects on cognitive function, general cognitive function, frontal lobe function, and attention/working memory were measured. Depression was evaluated using Korean version of Geriatric Depression Scale. The outcome measurements were performed before and after the 8 weeks of intervention(2 days per week). Results: After 8 weeks, general cognitive function, frontal cognitive function, attention/working memory function, depression of the dual-task training group were significantly increased than those of the single-task training group(p<0.05). Conclusion: The results indicated that an exercise-cognitive combined dual-task training for MCI was effective in improving general cognitive function, frontal /executive function, attention/working memory function and reducing depression.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4439-4455
• /
• 2021

Compared with two-stage object detection algorithms, one-stage algorithms provide a better trade-off between real-time performance and accuracy. However, these methods treat the intermediate features equally, which lacks the flexibility to emphasize meaningful information for classification and location. Besides, they ignore the interaction of contextual information from different scales, which is important for medium and small objects detection. To tackle these problems, we propose an image pyramid network based on dual attention mechanism (DAIPNet), which builds an image pyramid to enrich the spatial information while emphasizing multi-scale informative features based on dual attention mechanisms for one-stage object detection. Our framework utilizes a pre-trained backbone as standard detection network, where the designed image pyramid network (IPN) is used as auxiliary network to provide complementary information. Here, the dual attention mechanism is composed of the adaptive feature fusion module (AFFM) and the progressive attention fusion module (PAFM). AFFM is designed to automatically pay attention to the feature maps with different importance from the backbone and auxiliary network, while PAFM is utilized to adaptively learn the channel attentive information in the context transfer process. Furthermore, in the IPN, we build an image pyramid to extract scale-wise features from downsampled images of different scales, where the features are further fused at different states to enrich scale-wise information and learn more comprehensive feature representations. Experimental results are shown on MS COCO dataset. Our proposed detector with a 300 × 300 input achieves superior performance of 32.6% mAP on the MS COCO test-dev compared with state-of-the-art methods.

• Journal of IKEEE
• /
• v.27 no.3
• /
• pp.251-257
• /
• 2023

In this paper, we propose a dual-structured self-attention method that improves the lack of regional features of the vision transformer's self-attention. Vision Transformers, which are more computationally efficient than convolutional neural networks in object classification, object segmentation, and video image recognition, lack the ability to extract regional features relatively. To solve this problem, many studies are conducted based on Windows or Shift Windows, but these methods weaken the advantages of self-attention-based transformers by increasing computational complexity using multiple levels of encoders. This paper proposes a dual-structure self-attention using self-attention and neighborhood network to improve locality inductive bias compared to the existing method. The neighborhood network for extracting local context information provides a much simpler computational complexity than the window structure. CIFAR-10 and CIFAR-100 were used to compare the performance of the proposed dual-structure self-attention transformer and the existing transformer, and the experiment showed improvements of 0.63% and 1.57% in Top-1 accuracy, respectively.

• Annual Conference on Human and Language Technology
• /
• 2017.10a
• /
• pp.41-44
• /
• 2017

기계이해 시스템은 주어진 문서를 이해하고 질의에 해당하는 정답을 출력하는 방법으로 심층 신경망을 활용한 주의집중 방법이 발달하면서 활발히 연구되기 시작했다. 본 논문에서는 어휘 정보를 통해 문서와 질의를 이해하는 어휘 이해 모델과 품사 등장 정보, 의존 구문 정보를 통해 문법적 이해를 하는 구문 이해 모델을 함께 사용하여 기계이해 질의응답을 하는 Dual Bi-Directional Attention Flow모델을 제안한다. 한국어로 구성된 18,863개 데이터에서 제안 모델은 어휘 이해 모델만 사용하는 Bi-Directional Attention Flow모델보다 높은 성능(Exact Match: 0.3529, F1-score: 0.6718)을 보였다.

• 한국어정보학회:학술대회논문집
• /
• 2017.10a
• /
• pp.41-44
• /
• 2017

기계이해 시스템은 주어진 문서를 이해하고 질의에 해당하는 정답을 출력하는 방법으로 심층 신경망을 활용한 주의집중 방법이 발달하면서 활발히 연구되기 시작했다. 본 논문에서는 어휘 정보를 통해 문서와 질의를 이해하는 어휘 이해 모델과 품사 등장 정보, 의존 구문 정보를 통해 문법적 이해를 하는 구문 이해 모델을 함께 사용하여 기계이해 질의응답을 하는 Dual Bi-Directional Attention Flow모델을 제안한다. 한국어로 구성된 18,863개 데이터에서 제안 모델은 어휘 이해 모델만 사용하는 Bi-Directional Attention Flow모델보다 높은 성능(Exact Match: 0.3529, F1-score: 0.6718)을 보였다.

• Journal of Korean Society for Quality Management
• /
• v.49 no.4
• /
• pp.569-579
• /
• 2021

Purpose: The purpose of this study is the time series analysis for predicting the yield of crops applicable to each farm using environmental variables measured by smart farms cultivating tomato. In addition, it is intended to confirm the influence of environmental variables using a deep learning model that can be explained to some extent. Methods: A time series analysis was performed to predict production using environmental variables measured at 75 smart farms cultivating tomato in two periods. An LSTM-based encoder-decoder model was used for cases of several farms with similar length. In particular, Dual Attention Mechanism was applied to use environmental variables as exogenous variables and to confirm their influence. Results: As a result of the analysis, Dual Attention LSTM with a window size of 12 weeks showed the best predictive power. It was verified that the environmental variables has a similar effect on prediction through wieghtss extracted from the prediction model, and it was also verified that the previous time point has a greater effect than the time point close to the prediction point. Conclusion: It is expected that it will be possible to attempt various crops as a model that can be explained by supplementing the shortcomings of general deep learning model.

• The Journal of Korean Physical Therapy
• /
• v.35 no.3
• /
• pp.64-70
• /
• 2023

Purpose: This study investigated the influence of attention-demanding tasks on gait and measured differences in the temporal, spatial and kinematic characteristics between young healthy adults and elderly healthy adults. Methods: We recruited 16 healthy young adults and 15 healthy elderly adults in this study. All participants performed two cognitive tasks: a subtraction dual-task (SDT) and working memory dual-task (WMDT) during gait plus one normal gait. Using the LEGSys+ system, knee and hip-joint kinematic data during stance and swing phase and spatiotemporal parameter data were assessed in this study. Results: In the elderly adult group, attention-demanding tasks with gait showed a significant decrease in hip-joint motion during the stance phase, compared to the normal gait. Step length, stride length and stride velocity of the elderly adult group were significantly decreased in WMDT gait compared to normal gait (p<0.05). In the young adult group, kinematic data did not show any significant difference. However, stride velocity and cadence during SDT and WMDT gaits were significantly decreased compared to those of normal gait (p<0.05). Conclusion: We determined that attention-demanding tasks during gait in elderly adults can induce decreased hip-joint motion during stance phase and decreased gait speed and stride length to maintain balance and prevent risk of falling. We believe that understanding the changes during gait in older ages, particularly during attention-demanding tasks, would be helpful for intervention strategies and improved risk assessment.

• The Journal of Korean Physical Therapy
• /
• v.29 no.6
• /
• pp.303-306
• /
• 2017

Purpose: Gait variability is defined as the intrinsic fluctuations which occur during continuous gait cycles. Increased gait variability is closely associated with increased fall risk in older adults. This study investigated the influence of attention-demanding tasks on gait variability in elderly healthy adults. Methods: We recruited 15 healthy elderly adults in this study. All participants performed two cognitive tasks: a subtraction dual-task (SDT) and working memory dual-task (WMDT) during gait plus one normal gait. Using the $LEGSys^+$ system, we measured the coefficient of variation (CV %=$100{\times}$[standard deviation/mean]) for participants' stride time, stride length, and stride velocity. Results: SDT gait showed significant increment of stride time variability compared with usual gait (p<0.05), however, stride length and velocity variability did not difference between SDT gait and usual gait (p>0.05). WMDT gait showed significant increment of stride time and velocity variability compared with usual gait (p<0.05). In addition, stride time variability during WMDT gait also significantly increased compared with SDT gait (p<0.05). Conclusion: We reported that SDT and WMDT gait can induce the increment of the gait variability in elderly adults. We assume that attention demanding task based on working memory has the most influence on the interference between cognitive and gait function. Understanding the changes during dual task gait in older ages would be helpful for physical intervention strategies and improved risk assessment.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.496-498
• /
• 2023

Brain tumor segmentation problem has challenges in the tumor diversity of location, imbalance, and morphology. Attention mechanisms have recently been used widely to tackle medical segmentation problems efficiently by focusing on essential regions. In contrast, the fusion approaches enhance performance by merging mutual benefits from many models. In this study, we proposed a 3D dual fusion attention network to combine the advantages of fusion approaches and attention mechanisms by residual self-attention and local blocks. Compared to fusion approaches and related works, our proposed method has shown promising results on the BraTS 2018 dataset.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.1
• /
• pp.16-37
• /
• 2022

Accurate liver segment segmentation based on radiological images is indispensable for the preoperative analysis of liver tumor resection surgery. However, most of the existing segmentation methods are not feasible to be used directly for this task due to the challenge of exact edge prediction with some tiny and slender vessels as its clinical segmentation criterion. To address this problem, we propose a novel deep learning based segmentation model, called Boundary-Aware Dual Attention Liver Segment Segmentation Model (BADA). This model can improve the segmentation accuracy of liver segments with enhancing the edges including the vessels serving as segment boundaries. In our model, the dual gated attention is proposed, which composes of a spatial attention module and a semantic attention module. The spatial attention module enhances the weights of key edge regions by concerning about the salient intensity changes, while the semantic attention amplifies the contribution of filters that can extract more discriminative feature information by weighting the significant convolution channels. Simultaneously, we build a dataset of liver segments including 59 clinic cases with dynamically contrast enhanced MRI(Magnetic Resonance Imaging) of portal vein stage, which annotated by several professional radiologists. Comparing with several state-of-the-art methods and baseline segmentation methods, we achieve the best results on this clinic liver segment segmentation dataset, where Mean Dice, Mean Sensitivity and Mean Positive Predicted Value reach 89.01%, 87.71% and 90.67%, respectively.