http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Human Activity Recognition with LSTM Using the Egocentric Coordinate System Key Points
Wesonga, Sheilla,Park, Jang-Sik The Korean Society of Industry Convergence 2021 한국산업융합학회 논문집 Vol.24 No.6
As technology advances, there is increasing need for research in different fields where this technology is applied. On of the most researched topic in computer vision is Human activity recognition (HAR), which has widely been implemented in various fields which include healthcare, video surveillance and education. We therefore present in this paper a human activity recognition system based on scale and rotation while employing the Kinect depth sensors to obtain the human skeleton joints. In contrast to previous approaches that use joint angles, in this paper we propose that each limb has an angle with the X, Y, Z axes which we employ as feature vectors. The use of the joint angles makes our system scale invariant. We further calculate the body relative direction in the egocentric coordinates in order to provide the rotation invariance. For the system parameters, we employ 8 limbs with their corresponding angles each having the X, Y, Z axes from the coordinate system as feature vectors. The extracted features are finally trained and tested with the Long short term memory (LSTM) Network which gives us an average accuracy of 98.3%.
( Wesonga Sheilla ),이상협,박장식 한국품질경영학회 2023 한국품질경영학회 학술대회 Vol.2023 No.0
본 논문에서는 무선신호(RF, Radio Frequency) 및 데이터를 활용하여 사람의 자세를 추정하고 이상행동을 감지한다. 자세추정과 이상행동 감지를 위하여 LSTM 및 GRU 모델을 비교 분석하고, 유전 알고리즘(GA, Genetic Algorithm)으로 자세추정 딥러닝 모델 최적화 방안을 제안한다. RF는 밀리미터파(mmWave)로 60GHz에서 64GHz 대역의 주파수이다. mmWave 데이터를 활용함으로써 CCTV 카메라 영상에서 발생할 수 있는 개인정보 침해 문제를 해소할 수 있으며, 커튼 등을 투과할 수 있는 특징을 활용할 수 있다. 딥러닝 모델은 입력 데이터의 통계적인 특성과 모델의 구조에 따라 성능의 변화가 발생할 수 있다. 본 논문에서는 입력 데이터에 대한 최적의 자세추정 딥러닝 모델의 구조를 GA를 활용하여 최적화하는 방안을 제안한다. 이상행동 감지는 키포인트의 좌표를 시퀀스 데이터로 변환하여 이를 순환 신경망, 즉 LSTM 및 GRU을 통해 행동인식을 수행한다. 학습 및 성능 시험을 위하여 무인시설에서 사람의 쓰러짐, 기물파손 행위에 대한 mmWave 포인트 클라우드 데이터와 RGB-depth 카마레의 3D 키포인트 수집하였다. 또한 정확한 성능 분석을 위아여 10가지 재활운동 데이터인 MARS( mmwave-based Assistive Rehabilitation System for Smart Healthcare) 데이터셋 활용하여 자세추정 및 행동인식 성능을 비교 분석하였다. 시뮬레이션 결과 종래의 자세추정 모델에 비해 오차가 개선되었으며, 사람 행동인식 성능이 향상됨을 확인하였다.
Performance Comparison of Human Activity Recognition for Unmanned Retails
Sheilla Wesonga,Nusrat Jahan Tahira,Jang-Sik Park 제어로봇시스템학회 2022 제어로봇시스템학회 국제학술대회 논문집 Vol.2022 No.11
Lately, the broad usage of technology in almost all aspects of life has led to the increase in research supporting technology advancement. One of these research topics is Human Activity Recognition (HAR) with diverse applicability which include and not limited to video surveillance, healthcare and education. In this paper, we present a study based on human activity recognition while employing the Kinect RGB and Depth sensor camera to recognize seven different human activities (7 classes). The joint angles extracted from the Kinect depth sensor each has 3 axes (X, Y, Z) for the 8 limbs employed in our experiment as the feature vectors. For the purpose of classifying the human activities, we train and test with 3 different state of the art recurrent neural network models (GRU, LSTM, Bi-LSTM). The comparison of the 3 recurrent neural network models shows that LSTM has a higher human activity classification accuracy at 96% and using the confusion matrix as the performance metric for all the models, we show classification per activity.
Mubarak Al-Shukeili,Ronald Wesonga The Korean Statistical Society 2023 Communications for statistical applications and me Vol.30 No.3
This study proposes a modification to the objective function of the support vector machine for the linearly non-separable case of a binary classifier y<sub>i</sub> ∈ {-1, 1}. The modification takes into account the position of each data item x<sub>i</sub> from its corresponding class centroid. The resulting optimization function involves the centroid mean vector, and the spread of data besides the support vectors, which should be minimized by the choice of hyper-plane β. Theoretical assumptions have been tested to derive an optimal separable hyperplane that yields the minimal misclassification rate. The proposed method has been evaluated using simulation studies and real-life COVID-19 patient outcome hospitalization data. Results show that the proposed method performs better than the classical linear SVM classifier as the sample size increases and is preferred in the presence of correlations among predictors as well as among extreme values.
( Sang-hyeop Lee ),( Sheilla Wesonga ),( Jang-Sik Park ) 한국산업융합학회 2022 한국산업융합학회 논문집 Vol.25 No.2
Prognostics and health management (PHM) is recently converging throughout the industry, one of the trending issue is to detect abnormal conditions at decanter centrifuge during water treatment facilities. Wastewater treatment operation produces corrosive gas which results failures on attached sensors. This scenario causes frequent sensor replacement and requires highly qualified manager’s visual inspection while replacing important parts such as bearings and screws. In this paper, we propose anomaly detection by measuring the vibration of the decanter centrifuge based on the video camera images. Measuring the vibration of the screw decanter by applying the optical flow technique, the amount of movement change of the corresponding pixel is measured and fed into the LSTM model. As a result, it is possible to detect the normal/warning/dangerous state based on LSTM classification. In the future work, we aim to gather more abnormal data in order to increase the further accuracy so that it can be utilized in the field of industry.
Defect Diagnosis and Classification of Machine Parts Based on Deep Learning
( Hyun-tae Kim ),( Sang-hyeop Lee ),( Sheilla Wesonga ),( Jang-sik Park ) 한국산업융합학회 2022 한국산업융합학회 논문집 Vol.25 No.2
The automatic defect sorting function of machinery parts is being introduced to the automation of the manufacturing process. In the final stage of automation of the manufacturing process, it is necessary to apply computer vision rather than human visual judgment to determine whether there is a defect. In this paper, we introduce a deep learning method to improve the classification performance of typical mechanical parts, such as welding parts, galvanized round plugs, and electro galvanized nuts, based on the results of experiments. In the case of poor welding, the method to further increase the depth of layer of the basic deep learning model was effective, and in the case of a circular plug, the surrounding data outside the defective target area affected it, so it could be solved through an appropriate pre-processing technique. Finally, in the case of a nut plated with zinc, since it receives data from multiple cameras due to its three-dimensional structure, it is greatly affected by lighting and has a problem in that it also affects the background image. To solve this problem, methods such as two-dimensional connectivity were applied in the object segmentation preprocessing process. Although the experiments suggested that the proposed methods are effective, most of the provided good/defective images data sets are relatively small, which may cause a learning balance problem of the deep learning model, so we plan to secure more data in the future.