• Journal of Information Processing Systems
• /
• v.10 no.3
• /
• pp.335-354
• /
• 2014

The accuracy of training-based activity recognition depends on the training procedure and the extent to which the training dataset comprehensively represents the activity and its varieties. Additionally, training incurs substantial cost and effort in the process of collecting training data. To address these limitations, we have developed a training-free activity recognition approach based on a fuzzy logic algorithm that utilizes a generic activity model and an associated activity semantic knowledge. The approach is validated through experimentation with real activity datasets. Results show that the fuzzy logic based algorithms exhibit comparable or better accuracy than other training-based approaches.

• ETRI Journal
• /
• v.40 no.4
• /
• pp.499-510
• /
• 2018

A vector space based augmented structural kinematic (VSASK) feature descriptor is proposed for human activity recognition. An action descriptor is built by integrating the structural and kinematic properties of the actor using vector space based augmented matrix representation. Using the local or global information separately may not provide sufficient action characteristics. The proposed action descriptor combines both the local (pose) and global (position and velocity) features using augmented matrix schema and thereby increases the robustness of the descriptor. A multiclass support vector machine (SVM) is used to learn each action descriptor for the corresponding activity classification and understanding. The performance of the proposed descriptor is experimentally analyzed using the Weizmann and KTH datasets. The average recognition rate for the Weizmann and KTH datasets is 100% and 99.89%, respectively. The computational time for the proposed descriptor learning is 0.003 seconds, which is an improvement of approximately 1.4% over the existing methods.

• Journal of Korea Multimedia Society
• /
• v.23 no.12
• /
• pp.1540-1551
• /
• 2020

Aiming at the problem that the existing human behavior recognition algorithm cannot fully utilize the multi-level spatio-temporal information of the network, a human behavior recognition algorithm based on a dense three-dimensional residual network is proposed. First, the proposed algorithm uses a dense block of three-dimensional residuals as the basic module of the network. The module extracts the hierarchical features of human behavior through densely connected convolutional layers; Secondly, the local feature aggregation adaptive method is used to learn the local dense features of human behavior; Then, the residual connection module is applied to promote the flow of feature information and reduced the difficulty of training; Finally, the multi-layer local feature extraction of the network is realized by cascading multiple three-dimensional residual dense blocks, and use the global feature aggregation adaptive method to learn the features of all network layers to realize human behavior recognition. A large number of experimental results on benchmark datasets KTH show that the recognition rate (top-l accuracy) of the proposed algorithm reaches 93.52%. Compared with the three-dimensional convolutional neural network (C3D) algorithm, it has improved by 3.93 percentage points. The proposed algorithm framework has good robustness and transfer learning ability, and can effectively handle a variety of video behavior recognition tasks.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.877-885
• /
• 2022

Vision-based activity recognition has been widely attempted at construction sites to estimate productivity and enhance workers' health and safety. Previous studies have focused on extracting an individual worker's postural information from sequential image frames for activity recognition. However, various trades of workers perform different tasks with similar postural patterns, which degrades the performance of activity recognition based on postural information. To this end, this research exploited a concept of human-object interaction, the interaction between a worker and their surrounding objects, considering the fact that trade workers interact with a specific object (e.g., working tools or construction materials) relevant to their trades. This research developed an approach to understand the context from sequential image frames based on four features: posture, object, spatial features, and temporal feature. Both posture and object features were used to analyze the interaction between the worker and the target object, and the other two features were used to detect movements from the entire region of image frames in both temporal and spatial domains. The developed approach used convolutional neural networks (CNN) for feature extractors and activity classifiers and long short-term memory (LSTM) was also used as an activity classifier. The developed approach provided an average accuracy of 85.96% for classifying 12 target construction tasks performed by two trades of workers, which was higher than two benchmark models. This experimental result indicated that integrating a concept of the human-object interaction offers great benefits in activity recognition when various trade workers coexist in a scene.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.499-503
• /
• 2008

In this paper, a novel human activity recognition method is proposed which utilizes independent components of activity shape information from image sequences and Hidden Markov Model (HMM) for recognition. Activities are represented by feature vectors from Independent Component Analysis (ICA) on video images, and based on these features; recognition is achieved by trained HMMs of activities. Our recognition performance has been compared to the conventional method where Principle Component Analysis (PCA) is typically used to derive activity shape features. Our results show that superior recognition is achieved with our proposed method especially for activities (e.g., skipping) that cannot be easily recognized by the conventional method.

• Journal of the Korean Society of Industry Convergence
• /
• v.24 no.6_1
• /
• pp.693-698
• /
• 2021

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%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.1
• /
• pp.321-344
• /
• 2017

Smart homes are the new generation of homes where pervasive computing is employed to make the lives of the residents more convenient. Human activity recognition (HAR) is a fundamental task in these environments. Since critical decisions will be made based on HAR results, accurate recognition of human activities with low uncertainty is of crucial importance. In this paper, a novel HAR method based on a difference of convex programming (DCP) problem is represented, which manages to handle uncertainty. For this purpose, given an input sensor data stream, a primary belief in each activity is calculated for the sensor events. Since the primary beliefs are calculated based on some abstractions, they naturally bear an amount of uncertainty. To mitigate the effect of the uncertainty, a DCP problem is defined and solved to yield secondary beliefs. In this procedure, the uncertainty stemming from a sensor event is alleviated by its neighboring sensor events in the input stream. The final activity inference is based on the secondary beliefs. The proposed method is evaluated using a well-known and publicly available dataset. It is compared to four HAR schemes, which are based on temporal probabilistic graphical models, and a convex optimization-based HAR procedure, as benchmarks. The proposed method outperforms the benchmarks, having an acceptable accuracy of 82.61%, and an average F-measure of 82.3%.

• Journal of Korea Society of Industrial Information Systems
• /
• v.16 no.4
• /
• pp.9-18
• /
• 2011

In this paper, we propose a method which recognizes human activity using model-based gaze direction estimation in an indoor environment. The method consists of two steps. First, we detect a head region and estimate its gaze direction as prior information in the human activity recognition. We use color and shape information for the detection of head region and use Bayesian Network model representing relationships between a head and a face for the estimation of gaze direction. Second, we recognize event and scenario describing the human activity. We use change of human state for the event recognition and use a rule-based method with combination of events and some constraints. We define 4 types of scenarios related to the gaze direction. We show performance of the gaze direction estimation and human activity recognition with results of experiments.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.3
• /
• pp.1189-1204
• /
• 2018

Detecting and capturing 3D human structures from the intensity-based image sequences is an inherently arguable problem, which attracted attention of several researchers especially in real-time activity recognition (Real-AR). These Real-AR systems have been significantly enhanced by using depth intensity sensors that gives maximum information, in spite of the fact that conventional Real-AR systems are using RGB video sensors. This study proposed a depth-based routine-logging Real-AR system to identify the daily human activity routines and to make these surroundings an intelligent living space. Our real-time routine-logging Real-AR system is categorized into two categories. The data collection with the use of a depth camera, feature extraction based on joint information and training/recognition of each activity. In-addition, the recognition mechanism locates, and pinpoints the learned activities and induces routine-logs. The evaluation applied on the depth datasets (self-annotated and MSRAction3D datasets) demonstrated that proposed system can achieve better recognition rates and robust as compare to state-of-the-art methods. Our Real-AR should be feasibly accessible and permanently used in behavior monitoring applications, humanoid-robot systems and e-medical therapy systems.

• Journal of Internet Computing and Services
• /
• v.21 no.4
• /
• pp.127-135
• /
• 2020

The purpose of this study is to develop a human activity recognition (HAR) system, which distinguishes 13 activities, including five activities commonly dealt with in conventional HAR researches and eight activities from the Eastern-Asian culture. The eight special activities include floor-sitting/standing, chair-sitting/standing, floor-lying/up, and bed-lying/up. We used a 3-axis accelerometer sensor on the wrist for data collection and designed a machine learning model for the activity classification. Data clustering through preprocessing and feature extraction/reduction is performed. We then tested six machine learning algorithms for recognition accuracy comparison. As a result, we have achieved an average accuracy of 99.7% for the 13 activities. This result is far better than the average accuracy of current HAR researches based on a smartwatch (89.4%). The superiority of the HAR system developed in this study is proven because we have achieved 98.7% accuracy with publically available 'pamap2' dataset of 12 activities, whose conventionally met the best accuracy is 96.6%.