Improved ELBO-assisted Transformer for Skeleton-Based Action Recognition

Human action recognition is a research hotspot in the field of ambient intelligence, serving as a foundation for the Internet of Healthcare Things (IoHT) and smart home wellness, with extensive application value. Currently, most research primarily focuses on the performance of single-modal approach for human action recognition. However, single-modal data cannot adequately capture the action characteristics of the human body. For instance, camera data lacks detailed micro-motion information, while radar data lacks visual appearance information. This limitation makes human action recognition systems susceptible to complex covariate influences. To address this issue, this paper proposes a Multi-Modal Fusion Model, M2FM, to accurately recognize various complex human action from millimeter-wave radar signals and video data. In the millimeter-wave radar branch, a LFNet network is constructed to capture richly hierarchical human action representation by extracting micro-Doppler feature and cadence velocity feature simultaneously. Specially, a Linear Feedforward Neural Network (LFNN) module is designed for modeling global-local feature of human action. In the video branch, a lightweight Transformer-based video analysis and action recognition network, STL-Former, is developed. Specially, an Agent downSampling (AS) attention module and a Linear Maxpooling (LM) attention module are designed to achieve context-aware downsampling with global receptive field and spatial vector sequence downsampling, respectively. Additionally, a multi-scale Linear attention (Litner) module is designed for modeling the spatio-temporal information of human action. Finally, the outputs from two branches are fused by a filter-based Dempster-Shafer theory. The experimental results on a self-collected Multi-modal Human Action Dataset, JH-MHAD, show that the M2FM model outperforms other advanced models, with an average recognition accuracy of 99.3% for eight different human actions, and it demonstrates strong adaptability under different conditions.

Representation of human actions as a sequence of human body movements or action attributes enables the development of models for human activity recognition and summarization. We present an extension of the low-rank representation (LRR) model, termed the clustering-aware structure-constrained low-rank representation (CS-LRR) model, for unsupervised learning of human action attributes from video data. Our model is based on the union-of-subspaces (UoS) framework, and integrates spectral clustering into the LRR optimization problem for better subspace clustering results. We lay out an efficient linear alternating direction method to solve the CS-LRR optimization problem. We also introduce a hierarchical subspace clustering approach, termed hierarchical CS-LRR, to learn the attributes without the need for a priori specification of their number. By visualizing and labeling these action attributes, the hierarchical model can be used to semantically summarize long video sequences of human actions at multiple resolutions. A human action or activity can also be uniquely represented as a sequence of transitions from one action attribute to another, which can then be used for human action recognition. We demonstrate the effectiveness of the proposed model for semantic summarization and action recognition through comprehensive experiments on five real-world human action datasets.

Human action recognition is a fundamental research problem in computer vision. The accuracy of human action recognition has important applications. In this book chapter, the authors use a YOLOv7-based model for human action recognition. To evaluate the performance of the model, the action recognition results of YOLOv7 were compared with those using CNN+LSTM, YOLOv5, and YOLOv4. Furthermore, a small human action dataset suitable for YOLO model training is designed. This data set is composed of images extracted from KTH, Weizmann, MSR data sets. In this book chapter, the authors make use of this data set to verify the experimental results. The final experimental results show that using the YOLOv7 model for human action recognition is very convenient and effective, compared with the previous YOLO model.

The intricacies of human motion and action offer a rich vein of information crucial for enhancing the interaction between humans and machines. Human action understanding serves as the linchpin in this dynamic, bridging advanced computational methodologies with real-world applications across various sectors. This dissertation delves into the nuanced realm of action recognition and pose estimation, underscoring the transition from theoretical frameworks to practical applications. Central to the presented research is the development of the Graph Convolution Network (GCN)-based InfoGCN model for human action recognition, which introduces an informa tion bottleneck principle to optimize learning efficiency and accuracy. Additionally, the Pose Relation Transformer (PORT) is crafted to address challenges in hand and human pose esti mation, particularly under conditions of occlusion, by employing strategies akin to Masked Language Modeling in natural language processing. Further advancements are manifested in InfoGCN++, designed to facilitate real-time ac tion recognition, thereby catering to immediate analytical needs in dynamic environments. Concurrently, CARING-AI emerges as a novel framework for generating context-aware hu man motion, capturing both temporal and spatial dimensions to produce more realistic and accurate motion representations. Anticipating future needs, the dissertation proposes CARING-AI++, focusing on the enhancement of hand-object interaction within human motion generation. This progression signifies a deliberate move towards solving user-specific problems, illustrating a profound understanding of the necessary integration of technology and practicality. The collective work outlined in this dissertation not only propels the understanding of human action but also crystallizes its application in real-world scenarios, thereby bridging the gap between theoretical research and practical utility. With implications for healthcare, autonomous vehicles, and interactive technologies, the research underscores the potential of these advancements to revolutionize how we interact with and leverage technology for societal benefit, paving the way for future interdisciplinary collaboration and the ethical use of action recognition systems.

Scene image and human skeleton-based dual-stream human action recognition

Computing devices that can recognize various human activities or movements can be used to assist people in healthcare, sports, or human–robot interaction. Readily available data for this purpose can be obtained from the accelerometer and the gyroscope built into everyday smartphones. Effective classification of real-time activity data is, therefore, actively pursued using various machine learning methods. In this study, the transformer model, a deep learning neural network model developed primarily for the natural language processing and vision tasks, was adapted for a time-series analysis of motion signals. The self-attention mechanism inherent in the transformer, which expresses individual dependencies between signal values within a time series, can match the performance of state-of-the-art convolutional neural networks with long short-term memory. The performance of the proposed adapted transformer method was tested on the largest available public dataset of smartphone motion sensor data covering a wide range of activities, and obtained an average identification accuracy of 99.2% as compared with 89.67% achieved on the same data by a conventional machine learning method. The results suggest the expected future relevance of the transformer model for human activity recognition.

Bidirectional Long Short-Term Memory with Temporal Dense Sampling for human action recognition

Clinicians lack objective means for monitoring if their knee osteoarthritis patients are improving outside of the clinic (e.g., at home). Previous human activity recognition (HAR) models using wearable sensor data have only used data from healthy people and such models are typically imprecise for people who have medical conditions affecting movement. HAR models designed for people with knee osteoarthritis have classified rehabilitation exercises but not the clinically relevant activities of transitioning from a chair, negotiating stairs and walking, which are commonly monitored for improvement during therapy for this condition. Therefore, it is unknown if a HAR model trained on data from people who have knee osteoarthritis can be accurate in classifying these three clinically relevant activities. Therefore, we collected inertial measurement unit (IMU) data from 18 participants with knee osteoarthritis and trained convolutional neural network models to identify chair, stairs and walking activities, and phases. The model accuracy was 85% at the first level of classification (activity), 89–97% at the second (direction of movement) and 60–67% at the third level (phase). This study is the first proof-of-concept that an accurate HAR system can be developed using IMU data from people with knee osteoarthritis to classify activities and phases of activities.

Given the prevalence of surveillance cameras in our daily lives, human action recognition from videos holds significant practical applications. A persistent challenge in this field is to develop more efficient models capable of real-time recognition with high accuracy for widespread implementation. In this research paper, we introduce a novel human action recognition model named Context-Aware Memory Attention Network (CAMA-Net), which eliminates the need for optical flow extraction and 3D convolution which are computationally intensive. By removing these components, CAMA-Net achieves superior efficiency compared to many existing approaches in terms of computation efficiency. A pivotal component of CAMA-Net is the Context-Aware Memory Attention Module, an attention module that computes the relevance score between key-value pairs obtained from the 2D ResNet backbone. This process establishes correspondences between video frames. To validate our method, we conduct experiments on four well-known action recognition datasets: ActivityNet, Diving48, HMDB51 and UCF101. The experimental results convincingly demonstrate the effectiveness of our proposed model, surpassing the performance of existing 2D-CNN based baseline models. Article Highlights Recent human action recognition models are not yet ready for practical applications due to high computation needs. We propose a 2D CNN-based human action recognition method to reduce the computation load. The proposed method achieves competitive performance compared to most SOTA 2D CNN-based methods on public datasets.

Recently, a significant amount of literature concerning machine learning techniques has focused on automatic recognition of activities performed by people. The main reason for this considerable interest is the increasing availability of devices able to acquire signals which, if properly processed, can provide information about human activities of daily living (ADL). The recognition of human activities is generally performed by machine learning techniques that process signals from wearable sensors and/or cameras appropriately arranged in the environment. Whatever the type of sensor, activities performed by human beings have a strong subjective characteristic that is related to different factors, such as age, gender, weight, height, physical abilities, and lifestyle. Personalization models have been studied to take into account these subjective factors and it has been demonstrated that using these models, the accuracy of machine learning algorithms can be improved. In this work we focus on the recognition of human activities using signals acquired by the accelerometer embedded in a smartphone. The contributions of this research are mainly three. A first contribution is the definition of a clear validation model that takes into account the problem of personalization and which thus makes it possible to objectively evaluate the performances of machine learning algorithms. A second contribution is the evaluation, on three different public datasets, of a personalization model which considers two aspects: the similarity between people related to physical aspects (age, weight, and height) and similarity related to intrinsic characteristics of the signals produced by these people when performing activities. A third and last contribution is the development of a personalization model that considers both the physical and signal similarities. The experiments show that the employment of personalization models improves, on average, the accuracy, thus confirming the soundness of the approach and paving the way for future investigations on this topic.

In this paper, we present an original and efficient method of human action recognition in a video sequence. The proposed model is based on the generation and fusion of a set of prototypes generated from different view-points of the data cube of the video sequence. More precisely, each prototype is generated by using a multidimensional scaling (MDS) based nonlinear dimensionality reduction technique both along the temporal axis but also along the spatial axis (row and column) of the binary video sequence of 2D silhouettes. This strategy aims at modeling each human action in a low dimensional space, as a trajectory of points or a specific curve, for each viewpoint of the video cube in a complementary way. A simple K-NN classifier is then used to classify the prototype, for a given viewpoint, associated with each action to be recognized and then the fusion of the classification results for each viewpoint allow us to significantly improve the recognition rate performance. The experiments of our approach have been conducted on the publicly available Weizmann data-set and show the sensitivity of the proposed recognition system to each individual viewpoint and the efficiency of our multi-viewpoint based fusion approach compared to the best existing state-of-the-art human action recognition methods recently proposed in the literature.

Human activity recognition (HAR) is one of the leading research fields in ubiquitous computing working to integrate seamless technologies in our daily lives. The researches in this field focus on the technological advancement of fine activity recognition through using minimal technological deployments along with consideration of human factors. The domain specific knowledge of human activity recognition has shed significant importance on the activity data processing, due to the innate random nature of the activity data. The necessity for improved prediction accuracy of activities within limited infrastructural support and cost is necessary for the research works to be conducted at large. The data set utilized in this study is modified through tree-based feature-engineering method, which has significantly impacted the intrinsic time-series pattern associated with the original data set acquired from UCI machine learning repository. In this paper, the primary focus lies on improving classification accuracy of the human activities included in the data set through utilizing deep learning classifier models. The particular modified pattern of the data devoid of time-series pattern to major extent has resulted in lower classification accuracy by benchmark time-series classification models namely Keras-LSTM and RNN-LSTM. CNN based networks have demonstrated consisted improved training performance on the data set. The state-of-the-art classifier ResNet introduced by Microsoft Research in 2015 has achieved 99.9% classification accuracy on the data set of this study. Another important aim of this study is to infer the risk attached with adversarial attack on machine learning models. The high vulnerability of machine learning models in face of adversarial attacks has emerged as security concern of the deployment pipeline of machine learning models. To the author’s best knowledge, there has been no prior study to find out the high insecurity associated with the human activity recognition data set and concerned benchmark models. In this study, two prevalent adversarial approaches namely Fast Gradient Sign Method (FGSM) and Basic Iterative Method (BIM) have been employed to carry out non-targeted attack on the pre-trained ResNet model. The accuracy loss by the adversarial attack over a small range of added perturbation is tremendous, nearly 96% accuracy is diminished in the process. Finally, we have demonstrated the comparative performance by FGSM and BIM on reducing the classification accuracy of pre-trained model. The findings shed light on the nature of the data influencing the classification accuracy and the resiliency of state-of-the-art classifier model on facing adversarial attacks.

The enormous applicability of advanced human-computer interaction technology in daily life has facilitated the interest of the researchers toward the development of more intelligent autonomous systems. These human-computer interaction systems can be successful in real-life by addressing the gap among the existing techniques. This research work focuses on one of the prominent human-computer interaction applications of human activity recognition. Human activity recognition (HAR) is the process to detect human gestures, actions, and different types of interactions. The HAR process requires competent knowledge about the day to day human activities and advanced technology to recognize their activities. The conventional pattern recognition techniques are significant to recognize the human activities using the machine learning techniques but only within the controlled environment for the recognition of limited actions. In recent years, deep learning techniques are developed that can learn the deep attributes of the problem application and determine the outcomes with promising performance. The present paper has presented a systematic review of the deep learning models for video-based human activity recognition. The work describes the recent developments in the field for the analysis of different models. The paper also discusses the process of human activity recognition and the eminent datasets available for experimentation. The summarization of the work is illustrated with the future directions in the field.KeywordsDeep learningHuman computer interactionHuman activity recognitionVision based human activity recognitionVideo surveillanceVideo dataset

The launching of low-cost depth sensors opens up new potentials for RGB-D-based human action recognition. However, most of current RGB-D-based methods simply fuse multimodal features in a holistic manner and ignore the latent connections among different modalities. In this paper, we propose a multimodal correlative representation learning (MCRL) model for human action recognition from RGB-D videos. Specifically, we propose a spatio-temporal pyramid Fourier HOG feature (STPF-HOG) to capture local dynamic patterns around each human joint, which integrates both spatial arrangement and temporal structures. The proposed MCRL model utilizes multimodal data (skeleton, depth, and RGB), and learns shared structures among different modalities. The original low-level features are effectively compressed and projected to a latent subspace. Then, the discriminative shared features are learned by a supervised fashion. We formulate both subspace learning and shared features mining in a modified multi-task learning framework and solve the formulation using an iterative optimization algorithm. To perform computationally efficient action recognition, a robust collaborative representation classifier is presented by introducing a weight regularization matrix. Experimental results on three action datasets demonstrate that the proposed method leads to a more favorable performance compared with the state-of-the-art methods.

Human action recognition is essential in many real-world scenarios, such as video surveillance, human–computer interaction, and behavior analysis. Despite the progress in deep learning, issues such as occlusion, distraction from the background, and motion pattern variability still exist, thus restricting the generalization ability of current models. Most methods are based only on spatial or temporal features and cannot efficiently capture both in one framework, causing lower accuracy in realistic situations. In response to these shortcomings, a multilevel feature extraction approach was proposed by integrating spatial and temporal features to improve the action recognition precision. The method captures RGB frames, optical flow, spatial saliency maps, and temporal saliency maps to enable an overall inspection of video streams. Efficient feature extraction was achieved by applying a pre-trained Inception V3 model and then bidirectional gated recurrent units (Bi-GRUs) to include sequential modeling. An attention mechanism was also included to boost the classification process by focusing on key temporal segments. UCF101 and HMDB51 benchmark datasets evaluated the efficiency of the strategy. The model’s accuracy was 98.13% on UCF101 and 81.45% on HMDB51, which validated the superior discrimination ability of the model in processing heterogeneous human actions. These results confirm that the provided framework is an efficient and discriminative action recognition approach, thus suitable for applications requiring extensive motion analysis and real-time deployment.