• 로봇학회논문지
• /
• 제19권2호
• /
• pp.196-202
• /
• 2024

The ability to understand given environments and plan a sequence of actions leading to goal state is crucial for personal service robots. With recent advancements in deep learning, numerous studies have proposed methods for state representation in planning. However, previous works lack explicit information about relationships between objects when the state observation is converted to a single visual embedding containing all state information. In this paper, we introduce graph-based state representation that incorporates both object and relationship features. To leverage these advantages in addressing the task planning problem, we propose a Graph Neural Network (GNN)-based subgoal prediction model. This model can extract rich information about object and their interconnected relationships from given state graph. Moreover, a search-based algorithm is integrated with pre-trained subgoal prediction model and state transition module to explore diverse states and find proper sequence of subgoals. The proposed method is trained with synthetic task dataset collected in simulation environment, demonstrating a higher success rate with fewer additional searches compared to baseline methods.

• International Journal of Control, Automation, and Systems
• /
• 제6권6호
• /
• pp.904-914
• /
• 2008

An action-selection-mechanism(ASM) has been proposed to work as a fully connected finite state machine to deal with sequential behaviors as well as to allow a state in the task program to migrate to any state in the task, in which a primitive node in association with a state and its transitional conditions can be easily inserted/deleted. Also, such a primitive node can be learned by a shortest path-finding-based reinforcement learning technique. Specifically, we define a behavioral motivation as having state-dependent value as a primitive node for action selection, and then sequentially construct a network of behavioral motivations in such a way that the value of a parent node is allowed to flow into a child node by a releasing mechanism. A vertical path in a network represents a behavioral sequence. Here, such a tree for our proposed ASM can be newly generated and/or updated whenever a new behavior sequence is learned. To show the validity of our proposed ASM, experimental results of a mobile robot performing the task of pushing- a- box-in to- a-goal(PBIG) will be illustrated.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2019년도 춘계학술발표대회
• /
• pp.472-474
• /
• 2019

Because of the difference in network structure and loss function, Verification and identification models have their respective advantages and limitations for person reidentification (re-ID). In this work, we propose a multi-task network simultaneously computes the identification loss and verification loss for person reidentification. Given a pair of images as network input, the multi-task network simultaneously outputs the identities of the two images and whether the images belong to the same identity. In experiments, we analyze the major factors affect the accuracy of person reidentification. To address the occlusion problem and improve the generalization ability of reID models, we use the Random Erasing Augmentation (REA) method to preprocess the images. The method can be easily applied to different pre-trained networks, such as ResNet and VGG. The experimental results on the Market1501 datasets show significant and consistent improvements over the state-of-the-art methods.

• Smart Structures and Systems
• /
• 제31권4호
• /
• pp.383-392
• /
• 2023

Post-earthquake building condition assessment is crucial for subsequent rescue and remediation and can be automated by emerging computer vision and deep learning technologies. This study is based on an endeavour for the 2nd International Competition of Structural Health Monitoring (IC-SHM 2021). The task package includes five image segmentation objectives - defects (crack/spall/rebar exposure), structural component, and damage state. The structural component and damage state tasks are identified as the priority that can form actionable decisions. A multi-task Convolutional Neural Network (CNN) is proposed to conduct the two major tasks simultaneously. The rest 3 sub-tasks (spall/crack/rebar exposure) were incorporated as auxiliary tasks. By synchronously learning defect information (spall/crack/rebar exposure), the multi-task CNN model outperforms the counterpart single-task models in recognizing structural components and estimating damage states. Particularly, the pixel-level damage state estimation witnesses a mIoU (mean intersection over union) improvement from 0.5855 to 0.6374. For the defect detection tasks, rebar exposure is omitted due to the extremely biased sample distribution. The segmentations of crack and spall are automated by single-task U-Net but with extra efforts to resample the provided data. The segmentation of small objects (spall and crack) benefits from the resampling method, with a substantial IoU increment of nearly 10%.

• 정보과학회 논문지
• /
• 제43권8호
• /
• pp.878-882
• /
• 2016

이미지의 내용을 설명하는 캡션을 자동으로 생성하는 기술은 이미지 인식과 자연어처리 기술을 필요로 하는 매우 어려운 기술이지만, 유아 교육이나 이미지 검색, 맹인들을 위한 네비게이션 등에 사용될 수 있는 중요한 기술이다. 본 논문에서는 이미지 캡션 생성을 위해 Convolutional Neural Network(CNN)으로 인코딩된 이미지 정보를 입력으로 갖는 이미지 캡션 생성에 최적화된 Recurrent Neural Network(RNN) 모델을 제안하고, 실험을 통해 본 논문에서 제안한 모델이 Flickr 8K와 Flickr 30K, MS COCO 데이터 셋에서 기존의 연구들보다 높은 성능을 얻음을 보인다.

• Management Science and Financial Engineering
• /
• 제17권2호
• /
• pp.39-62
• /
• 2011

In wireless sensor network, since each sensor is equipped with a limited power, efficient use of the energy is important. One possible network management scheme is to cluster the sensors into several sets, so that the sensors in each of the sets can completely perform the monitoring task. Then the sensors in one set become active to perform the monitoring task and the rest of the sensors switch to a sleep state to save energy. Therefore, we rotate the roles of the active set among the sensors to maximize the network lifetime. In this paper, we suggest an optimal algorithm for the maximum lifetime coverage problem which maximizes the network lifetime. For comparison, we implemented both the heuristic proposed earlier and our algorithm, and executed computational experiments. Our algorithm outperformed the heuristic concerning the obtained network lifetimes, and it found the solutions in a reasonable amount of time.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권4호
• /
• pp.1275-1292
• /
• 2021

As a fine-grained classification problem, aspect-level sentiment classification predicts the sentiment polarity for different aspects in context. To address this issue, researchers have widely used attention mechanisms to abstract the relationship between context and aspects. Still, it is difficult to effectively obtain a more profound semantic representation, and the strong correlation between local context features and the aspect-based sentiment is rarely considered. In this paper, a hybrid attention capsule network for aspect-level sentiment classification (ABASCap) was proposed. In this model, the multi-head self-attention was improved, and a context mask mechanism based on adjustable context window was proposed, so as to effectively obtain the internal association between aspects and context. Moreover, the dynamic routing algorithm and activation function in capsule network were optimized to meet the task requirements. Finally, sufficient experiments were conducted on three benchmark datasets in different domains. Compared with other baseline models, ABASCap achieved better classification results, and outperformed the state-of-the-art methods in this task after incorporating pre-training BERT.

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제9권11호
• /
• pp.265-272
• /
• 2020

뇌파를 이용한 Brain-computer interface (BCI) 연구에서는 다른 그룹보다 그 성능을 발휘하지 못하는 소위 BCI-illiteracy 그룹이라고 알려진 사용자 집단에 대한 이해와 처리가 중요하다. 본 연구는 사용자로부터 사전 휴지 상태의 뇌파 신호를 미리 측정하고 그 신호로부터 주파수 기반의 특징 변수를 생성하여 이를 피험자 개인의 특성 변수로 사용하고, 추정된 개인 특성 변수를 이용하여 이후 움직임 상상 패러다임이 적용된 BCI 시행의 성능과 어느 정도의 정량적 연관성을 가지며 이를 정확하게 예측할 수 있는지를 밝히고자 하였다. 결과에 대한 신뢰성을 높이기 위해서 검증된 공개 뇌파 데이터베이스를 활용하고 Convolution neural network 기반의 딥러닝 기법을 활용하여 이진 BCI 성능 계산을 실시하였으며 Lasso 정규화가 적용된 선형 회귀 분석을 통해서 각 특징 변수와의 예측 관련성을 조사하였다. 첫 번째로 휴지 상태 뇌파 모든 특징 변수들과 BCI 성능 간의 연관성을 파악하기 위해서 전통적인 통계 방법들을 적용하였고 이를 통해서 전두엽에서 측정된 뇌파 신호들의 13 Hz를 기준으로 이보다 낮은 주파수와 높은 주파수 파워 간의 비율이 BCI 성능 사이와 통계적 유의미한 높은 상관성이 가지고 있다는 사실을 확인할 수 있었다. 이를 근거로 상대 주파수 비율 값이 BCI 성능을 예측해볼 수 있는 좋은 지표 후보군으로 지정하였다. 두 번째로 Lasso를 이용한 회귀 분석을 통해서 휴식 상태의 상대 주파수 비율 변수를 이용하여 BCI 성능 사이에 최대 선형 계수 0.544 수준의 선형 관계를 찾을 수 있었으며, BCI 과제를 잘 시행할 수 있는 그룹과 못할 그룹을 AUC 0.817 수준으로 예측할 수 있었다. 본 연구에서는 각 사용자마다 측정된 휴지 상태의 뇌파로부터 앞으로 있을 BCI 성능을 예측할 수 있는 방법론 제시함으로써 일반인을 대상으로 좀 더 신뢰성 있고 응용 가능한 BCI 시스템 개발에 기여하고자 한다.

• 제어로봇시스템학회논문지
• /
• 제11권2호
• /
• pp.103-108
• /
• 2005

Loading/Unloading task in the real industry is performed by crane, but most of the loading/unloading task with the weight of 5kg∼30kg is done by human workers and this kind of work causes industrial disaster of workers. Therefore it is necessary to develop low cost loading/unloading manipulator system to prevent this kind of industrial accidents. This paper is concerned with the design and fabrication of 2 axis pneumatic manipulators using on/off solenoid valves and accurate position control without respect to the external load and low damping in the pneumatic rotary actuator. To overcome the change of external load, switching of control parameter using LVQNN (Learning Vector Quantization Neural Network) is newly applied, which estimates the external loads in the pneumatic cylinder. As an underlying controller, a state feedback controller using position, velocity and acceleration is applied to the switching control system. The effectiveness of the proposed control algorithms are demonstrated through experiments of pneumatic cylinder with various loads.

• International Journal of Computer Science & Network Security
• /
• 제22권3호
• /
• pp.163-172
• /
• 2022

The interconnection of an enormous number of devices into the Internet at a massive scale is a consequence of the Internet of Things (IoT). As a result, tasks offloading from these IoT devices to remote cloud data centers become expensive and inefficient as their number and amount of its emitted data increase exponentially. It is also a challenge to optimize IoT device energy consumption while meeting its application time deadline and data delivery constraints. Consequently, Fog Computing was proposed to support efficient IoT tasks processing as it has a feature of lower service delay, being adjacent to IoT nodes. However, cloud task offloading is still performed frequently as Fog computing has less resources compared to remote cloud. Thus, optimized schemes are required to correctly characterize and distribute IoT devices tasks offloading in a hybrid IoT, Fog, and cloud paradigm. In this paper, we present a detailed survey and classification of of recently published research articles that address the energy efficiency of task offloading schemes in IoT-Fog-Cloud paradigm. Moreover, we also developed a taxonomy for the classification of these schemes and provided a comparative study of different schemes: by identifying achieved advantage and disadvantage of each scheme, as well its related drawbacks and limitations. Moreover, we also state open research issues in the development of energy efficient, scalable, optimized task offloading schemes for Fog computing.