• ETRI Journal
• /
• 제42권1호
• /
• pp.78-89
• /
• 2020

Human activity recognition (HAR) has become effective as a computer vision tool for video surveillance systems. In this paper, a novel biometric system that can detect human activities in 3D space is proposed. In order to implement HAR, joint angles obtained using an RGB-depth sensor are used as features. Because HAR is operated in the time domain, angle information is stored using the sliding kernel method. Haar-wavelet transform (HWT) is applied to preserve the information of the features before reducing the data dimension. Dimension reduction using an averaging algorithm is also applied to decrease the computational cost, which provides faster performance while maintaining high accuracy. Before the classification, a proposed thresholding method with inverse HWT is conducted to extract the final feature set. Finally, the K-nearest neighbor (k-NN) algorithm is used to recognize the activity with respect to the given data. The method compares favorably with the results using other machine learning algorithms.

• 한국멀티미디어학회논문지
• /
• 제15권4호
• /
• pp.417-424
• /
• 2012

In this paper, we propose an automatic system that recognizes dynamic human gestures activity, including Arabic numbers from 0 to 9. We assume the gesture trajectory is almost in a plane that called principal gesture plane, then the Least Squares Method is used to estimate the plane and project the 3-D trajectory model onto the principal. An improved FNN model combined with HMM is proposed for dynamic gesture recognition, which combines ability of HMM model for temporal data modeling with that of fuzzy neural network. The proposed algorithm shows that satisfactory performance and high recognition rate.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권5호
• /
• pp.2327-2347
• /
• 2018

Human actions can be recognized from depth sequences. In the proposed algorithm, we initially construct depth, motion maps (DMM) by projecting each depth frame onto three orthogonal Cartesian planes and add the motion energy for each view. The body part of the action (BPoA) is calculated by using bounding box with an optimal window size based on maximum spatial and temporal changes for each DMM. Furthermore, feature vector is constructed by using BPoA for each human action view. In this paper, we employed an ensemble based learning approach called Rotation Forest to recognize different actions Experimental results show that proposed method has significantly outperforms the state-of-the-art methods on Microsoft Research (MSR) Action 3D and MSR DailyActivity3D dataset.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권7호
• /
• pp.3599-3619
• /
• 2019

In human activity recognition system both static and motion information play crucial role for efficient and competitive results. Most of the existing methods are insufficient to extract video features and unable to investigate the level of contribution of both (Static and Motion) components. Our work highlights this problem and proposes Static-Motion fused features descriptor (SMFD), which intelligently leverages both static and motion features in the form of descriptor. First, static features are learned by two-stream 3D convolutional neural network. Second, trajectories are extracted by tracking key points and only those trajectories have been selected which are located in central region of the original video frame in order to to reduce irrelevant background trajectories as well computational complexity. Then, shape and motion descriptors are obtained along with key points by using SIFT flow. Next, cholesky transformation is introduced to fuse static and motion feature vectors to guarantee the equal contribution of all descriptors. Finally, Long Short-Term Memory (LSTM) network is utilized to discover long-term temporal dependencies and final prediction. To confirm the effectiveness of the proposed approach, extensive experiments have been conducted on three well-known datasets i.e. UCF101, HMDB51 and YouTube. Findings shows that the resulting recognition system is on par with state-of-the-art methods.

• 디지털콘텐츠학회 논문지
• /
• 제18권3호
• /
• pp.559-565
• /
• 2017

스마트폰에 내장된 가속도 센서를 이용하여 사용자의 동작 상태나 행동을 인식하기 위한 연구가 다양하게 진행되어 왔다. 본 논문에서는 스마트폰의 3D 가속도 정보에 신경회로망을 적용하여 사람의 행동을 인식하는 연구를 진행하였다. 시계열 데이터를 신경회로망에 그대로 적용하면 성능상의 문제가 발생한다. 따라서 여러 시구간에 대해 특징을 추출하여 각 시구간에 대해 신경회로망을 학습시키고, 이 신경회로망들의 출력들을 입력으로 하여 학습하여 구성하는 다중 시구간 신경회로망을 제안하였다. 제안하는 방법을 실제 가속도 데이터에 적용한 결과 SVM, AdaBoost, IBk 등 다른 분류기보다 우수한 성능을 보였다.

• Archives of Pharmacal Research
• /
• 제28권12호
• /
• pp.1317-1323
• /
• 2005

In an attempt to examine the ability of benzisothiazole-based drugs to interact with $\beta$-adrenoceptors, a series of 1,2-benzisothiazole derivatives, which were substituted with various propanolamine or oxypropanolamine side chains in the 2 or 3 position, were synthesised and tested. The pharmacological activity of these compounds at the ,$\beta$-adrenoceptors was examined using isolated rat atria and small intestinal segments, which preferentially express the $\beta_{1}$- and $\beta_{3}$-adrenoceptor-mediated responses, respectively. None of these products showed any $\beta$-adrenoceptor agonistic activity. In contrast, the 2- and 3-substituted isopropyl, tert-butyl, benzyl, and piperonyl derivatives 2a-d and 3a-d elicited surmountable inhibition of the isoprena­line-induced chronotropic effects in the atria, suggesting competitive antagonism at the $\beta_{1}$­recognition site. The $pA_{2}$ values revealed tert-butyl 3b and the isopropyl substituted piperonyl derivatives 3a to be the most effective. Remarkably, many of the 2-substituted propanolamines were less active than the corresponding 3-substituted oxypropanolamines. With the exception of compound 3b, none of these drugs antagonised the muscle relaxant activity of isoprenaline in the intestine, suggesting no effect on the $\beta_{3}$-adrenoceptors. These results confirm the ability of the benzisothiazole ring to interact with the $\beta$-adrenoceptors, and demonstrate that 2-substitution with propanolamine or 3-substitution with oxypropanolamine groups yields compounds with preferential antagonistic activity at the cardiac $\beta_{1}$adrenoceptors. The degree of antagonism depends strongly on both the nature of the substituent and its position on the benzisothiazole ring.

• 정보과학회 논문지
• /
• 제41권11호
• /
• pp.927-934
• /
• 2014

본 논문에서는 Kinect와 같은 RGB-D 센서를 이용하여 사람의 3차원 신체 포즈 스트림 데이터를 생성하고, 이로부터 사람의 일상 행위를 효과적으로 인식하는 방법을 제안한다. Kinect SDK나 OpenNI에서 제공하는 실시간 신체 포즈 데이터는 Kinect 중심의 3차원 데카르트 좌표계로 표현되기 때문에, 시점 변화 문제와 크기 변화 문제를 겪을 가능성이 높다. 이러한 문제를 해결하고 시점 및 크기 불변인 특징을 얻기 위해, 본 논문에서는 신체 포즈 데이터를 실험자의 골반을 원점으로 하는 구면 좌표계로 변환하고 실험자의 팔 길이를 이용한 크기 정규화를 수행한다. 또한, 본 논문에서는 확률 그래프 모델 중 하나인 은닉 조건부 랜덤 필드를 이용하여, 고수준의 일상 행위들이 내포하는 다양한 내부 구조를 효과적으로 표현한다. 두 가지 데이터 집합 KAD-70과 CAD-60을 이용한 실험을 통해, 본 논문에서 제안한 행위 인식 방법과 구현 시스템의 높은 인식 성능을 확인하였다.

• 정보처리학회논문지D
• /
• 제18D권2호
• /
• pp.143-148
• /
• 2011

본 논문에서는 3축 가속도 센서를 소형 디바이스(활동량 측정기)로 구성하고 이를 사람의 신체에 착용하고 사람이 보행 시 발생하는 가속도 센서의 Raw 데이터 출력 값을 획득하여 실시간 활동량으로 변환하고 모니터링 할 수 있는 활동량 측정기와 알고리즘을 개발하였다. 피험자 59명을 대상으로 트레드밀(Treadmill)에서 호흡가스대사분석기(K4B2), Actical 그리고 본 연구에서 개발된 활동량 측정기를 착용 후 36분 동안 테스트 프로토콜에 따라 다양한 속력의 걸음(느리게 걷기, 걷기, 빠르게 걷기, 천천히 뛰기, 뛰기, 빠르게 뛰기)에 대해서 실험을 하였다. 가속도 센서의 출력 데이터와 피험자 정보를 이용하여 에너지소비량(Energy Expenditure :EE)을 추정하는 회귀식을 도출하였으며 이는 실험시 같이 착용한 Actical보다 제안하는 활동량 변환 알고리즘의 성능이 1.61% 향상 되었다.

• Communications for Statistical Applications and Methods
• /
• 제30권6호
• /
• pp.551-560
• /
• 2023

Recently, there has been significant research into the recognition of human activities using three-dimensional sequential skeleton data captured by the Kinect depth sensor. Many of these studies employ deep learning models. This study introduces a novel feature selection method for this data and analyzes it using machine learning models. Due to the high-dimensional nature of the original Kinect data, effective feature extraction methods are required to address the classification challenge. In this research, we propose using the first four moments as predictors to represent the distribution of joint sequences and evaluate their effectiveness using two datasets: The exergame dataset, consisting of three activities, and the MSR daily activity dataset, composed of ten activities. The results show that the accuracy of our approach outperforms existing methods on average across different classifiers.

• Asian-Australasian Journal of Animal Sciences
• /
• 제26권5호
• /
• pp.638-645
• /
• 2013

Interferon-tau (IFNT) is thought to be the conceptus protein that signals maternal recognition of pregnancy in ruminants. We and others have observed that OCT4 expression persists in the trophectoderm of ruminants; thus, both CDX2 and OCT4 coexist during the early stages of conceptus development. The aim of this study was to examine the effect of CDX2 and OCT4 on IFNT gene transcription when evaluated with other transcription factors. Human choriocarcinoma JEG-3 cells were cotransfected with an ovine IFNT (-654-bp)-luciferase reporter (-654-IFNT-Luc) construct and several transcription factor expression plasmids. Cotransfection of the reporter construct with Cdx2, Ets2 and Jun increased transcription of -654-IFNT-Luc by about 12-fold compared with transfection of the construct alone. When cells were initially transfected with Oct4 (0 h) followed by transfection with Cdx2, Ets2 and/or Jun 24 h later, the expression of -654-IFNT-Luc was reduced to control levels. OCT4 also inhibited the stimulatory activity of CDX2 alone, but not when CDX2 was combined with JUN and/or ETS2. Thus, when combined with the other transcription factors, OCT4 exhibited little inhibitory activity towards CDX2. An inhibitor of the transcriptional coactivator CREB binding protein (CREBBP), 12S E1A, reduced CDX2/ETS2/JUN stimulated -654-IFNT-Luc expression by about 40%, indicating that the formation of an appropriate transcription factor complex is required for maximum expression. In conclusion, the presence of OCT4 may initially minimize IFNT expression; however, as elongation proceeds, the increasing expression of CDX2 and formation of the transcription complex leads to greatly increased IFNT expression, resulting in pregnancy establishment in ruminants.