Human Activity Analysis Research Articles

Monitoring and Analysis of Human Activities Using Heterogeneous Sensors in the Gas Room of a Hospital

Every hospital relies on the gas pipeline system. There are common three types of gases are used: oxygen, nitrogen, and nitrous oxide. Each type of gas is critical in many departments, including the operating room and intensive care unit. Gas room administrators, such as engineers or hospital personnel, are just as important as gas rooms. If something happens to engineers or personnel, an unattended gas room may malfunction and cause trouble in the hospital. Therefore, we proposed a monitoring system with a human activity surveillance project using accelerometer sensors, video images, and gas sensors. This system can collect all the signals used for the analysis of the posture or activities of the personnel who examine the gas in the gas room. We also developed surveillance in the event of an unexpected event in the gas room.

Sample size effects on landslide susceptibility models: A comparative study of heuristic, statistical, machine learning, deep learning and ensemble learning models with SHAP analysis

In landslide susceptibility assessment (LSA), inventory incompleteness impacts the accuracy of different models to varying degrees. However, this area remains under-researched. This study investigated six LSA models from heuristic, statistical, machine learning and ensemble learning models (analytical hierarchy process (AHP), frequency ratio (FR), logistic regression (LR), Keras based deep learning (KBDL), XGBoost, and LightGBM) across six different sample sizes (100%, 90%, 75%, 50%, 25%, and 10%). Results revealed that XGBoost and LightGBM consistently outperformed other models across all sample sizes. The LR and KBDL models followed, while FR model was the most affected by sample size variations. AHP, an empirical model, remained unaffected by sample size. Through SHapley Additive exPlanations (SHAP) analysis, elevation, NDVI, slope, land use, and distance to roads and rivers emerged as pivotal indicators for landslide occurrences in the study area, suggesting that human activities significantly influence these events. Five time-varying indicators regarding human activity and climate validated this inference, which provides a new method to identify landslide triggering factors, especially in areas of intense human activity. Based on the findings, a comprehensive framework for LSA is proposed to assist landslide managers in making informed decisions. Future research should focus on expanding model diversity to address the effects of sample size, enhancing the adaptability of the LSA framework, deepening the analysis of human activity impacts on landslides using explainable machine learning techniques, addressing temporal inventory incompleteness in LSA, and critically evaluating model sensitivity to sample size variations across multiple disciplines.

Versatile Graph Neural Networks Toward Intuitive Human Activity Understanding.

Benefiting from the advanced human visual system, humans naturally classify activities and predict motions in a short time. However, most existing computer vision studies consider those two tasks separately, resulting in an insufficient understanding of human actions. Moreover, the effects of view variations remain challenging for most existing skeleton-based methods, and the existing graph operators cannot fully explore multiscale relationship. In this article, a versatile graph-based model (Vers-GNN) is proposed to deal with those two tasks simultaneously. First, a skeleton representation self-regulated scheme is proposed. It is among the first trials that successfully integrate the idea of view adaptation into a graph-based human activity analysis system. Next, several novel graph operators are proposed to model the positional relationships and learn the abstract dynamics between different human joints and parts. Finally, a practical multitask learning framework and a multiobjective self-supervised learning scheme are proposed to promote both the tasks. The comparative experimental results show that Vers-GNN outperforms the recent state-of-the-art methods for both the tasks, with the to date highest recognition accuracies on the datasets of NTU RGB + D (CV: 97.2%), UWA3D (88.7%), and CMU (1000 ms: 1.13).

A human mesh-centered approach to action recognition in the operating room

Aim: Video review programs in hospitals play a crucial role in optimizing operating room workflows. In scenarios where split-seconds can change the outcome of a surgery, the potential of such programs to improve safety and efficiency is profound. However, leveraging this potential requires a systematic and automated analysis of human actions. Existing methods predominantly employ manual methods, which are labor-intensive, inconsistent, and difficult to scale. Here, we present an AI-based approach to systematically analyze the behavior and actions of individuals from operating rooms (OR) videos. Methods: We designed a novel framework for human mesh recovery from long-duration surgical videos by integrating existing human detection, tracking, and mesh recovery models. We then trained an action recognition model to predict surgical actions from the predicted temporal mesh sequences. To train and evaluate our approach, we annotated an in-house dataset of 864 five-second clips from simulated surgical videos with their corresponding actions. Results: Our best model achieves an F1 score and the area under the precision-recall curve (AUPRC) of 0.81 and 0.85, respectively, demonstrating that human mesh sequences can be successfully used to recover surgical actions from operating room videos. Model ablation studies suggest that action recognition performance is enhanced by composing human mesh representations with lower arm, pelvic, and cranial joints. Conclusion: Our work presents promising opportunities for OR video review programs to study human behavior in a systematic, scalable manner.

Deep Learning for Skeleton-Based Human Activity Segmentation: An Autoencoder Approach

Automatic segmentation is essential for enhancing human activity recognition, especially given the limitations of publicly available datasets that often lack diversity in daily activities. This study introduces a novel segmentation method that utilizes skeleton data for a more accurate and efficient analysis of human actions. By employing an autoencoder, this method extracts representative features and reconstructs the dataset, using the discrepancies between the original and reconstructed data to establish a segmentation threshold. This innovative approach allows for the automatic segmentation of activity datasets into distinct segments. Rigorous evaluations against ground truth across three publicly available datasets demonstrate the method’s effectiveness, achieving impressive average annotation error, precision, recall, and F1-score values of 3.6, 90%, 87%, and 88%, respectively. This illustrates the robustness of the proposed method in accurately identifying change points and segmenting continuous skeleton-based activities as compared to two other state-of-the-art techniques: one based on deep learning and another using the classical time-series segmentation algorithm. Additionally, the dynamic thresholding mechanism enhances the adaptability of the segmentation process to different activity dynamics improving overall segmentation accuracy. This performance highlights the potential of the proposed method to significantly advance the field of human activity recognition by improving the accuracy and efficiency of identifying and categorizing human movements.

Attribution Analysis of Climate Change and Human Activities on Runoff and Vegetation Changes in the Min River Basin

Hydrological processes and the sustainable use of water resources in a river basin are altered by climate change and changes in human variables. This study examined the significant effects of vegetation and hydrological, climatic, and human activity changes on the basin’s biological environment and usage of water resources. The Min River Basin (MRB) in the upper Yangtze River served as the study location. Mann–Kendall and Pettitt mutation test techniques were used to examine the features of runoff changes in the basin. The effects of meteorological and anthropogenic factors on runoff and vegetation changes in the MRB from 1982 to 2020 were quantitatively evaluated using the expanded Budyko equation. Following this, spatial and temporal variations in land use and the NDVI in the basin were studied. The results of the research demonstrated the following: (1) The MRB yearly runoff trended downward and that an abrupt change in runoff happened in 1994. (2) Precipitation (Pr) showed a decreasing tendency from the base period (S1) to the change period (S2), but potential evapotranspiration (ET0) showed an increasing trend. (3) From 1985 to 2020, the land use area of the MRB changed rapidly, and the construction land and water area increased by 322% and 58.85%, respectively, while the cultivated land area decreased by 11.72%. (4) From S1 to S2, there was a rising trend in both the NDVI and the Budyko parameter n. The contributions of Pr, ET0, NDVI, and n to the runoff change were 32.41%, 9.43%, 27.51%, and 30.65%, respectively.

Urbanization intensifies the imbalance between human development and biodiversity conservation: Insights from the coupling analysis of human activities and habitat quality

AbstractIntensified human activities have been seriously threatening the structure and ecological processes of ecosystems, resulting in habitat degradation. Therefore, coordinating the coupling between human activities and habitat quality (HQ) is crucial for high‐quality sustainable regional development and human well‐being. This study evaluated the human activities and HQ in the Pearl River Delta (PRD) urban agglomeration in China from 2000 to 2020 using the human footprint index (HFI) and the integrated valuation of ecosystem services and tradeoffs model. Then, we employed bivariate spatial autocorrelation and a coupling coordination degree (CCD) model to explore the synergistic relationship between human activities and HQ. The results show that spatial changes in HQ were predominantly driven by human activities. The gradual outward urban expansion resulted in significant HQ degradation. Slight HQ improvement by ecological restoration in urban outskirts cannot offset HQ losses caused by urbanization. During the study period, high‐HQ low‐HFI clusters decreased by 1.02%, while low‐HQ high‐HFI clusters increased by 4.67%, the two main clustering types in the PRD. Despite the CCD between HFI and HQ increased after 2010, the continuous changes of CCD characteristics from the HFI significantly lagged type to the HQ lagged. HFI showed an inverted U‐shaped relationship with CCD. The CCD peaks during 2000–2020 corresponded to HFI decreasing from 0.711 to 0.566. This indicates that the risk of decoupling between human activities and HQ gradually increased. Furthermore, CCD levels and characteristics in different bivariate clusters exhibited varying changes over time. These results reveal that the spatiotemporal dislocation between urbanization and ecological restoration induced the spatial nonstationarity of the coupling relationship between human activities and HQ. Urbanization exacerbates the imbalance between human development and biodiversity conservation. Therefore, we suggest reasonably delimiting urban boundaries, controlling the scale of urban sprawl, and strengthening biodiversity protection in areas undergoing rapid urbanization. In addition, we advocate for the division of ecological barrier zones, urban development buffer zones, and urban built‐up areas, each with tailored management and protection measures. Our findings can provide an important reference for the ecological restoration of urban agglomerations.

Impact Analysis of Human Activities on Mangrove Conservation in Coastal Sulawesi

Mangrove ecosystems are invaluable coastal habitats that provide essential ecological services and support diverse flora and fauna. However, human activities have increasingly threatened the sustainability of these ecosystems, particularly in regions like Coastal Sulawesi, Indonesia. This qualitative analysis investigates the impact of human activities on mangrove conservation in Coastal Sulawesi, employing semi-structured interviews, participant observation, and document analysis. The study reveals that drivers such as aquaculture expansion, unsustainable logging, pollution, and coastal development are contributing to mangrove degradation. Stakeholder perspectives vary, with local communities emphasizing the importance of mangroves for livelihoods and culture, while government officials stress the need for balancing conservation with economic development. Existing conservation efforts include protected area designation, community-based management projects, and awareness-raising campaigns. Collaborative action involving government, NGOs, academia, and local communities is essential to promote sustainable mangrove conservation in Coastal Sulawesi.

Analysis of Fitness Based on Smart Watch Data

Smart watch, a new trend to many young as well as aged persons now a days, can serve as a fitness tracker and be far more accurate than a phone. Besides many technological advantages, a smartwatch can easily collect the data related to fitness and other movements also. Smart watch provides such an organized human health and activities dataset so that by this dataset a human activities analysis can be performed and by this analysis a conclusion can be drawn about their health. Accordingly, our Project aim is to analyze the data and examine whether the data collected through smart watches can describe the physical fitness of individuals on the basis of some of the factors related to daily activities. Keywords— Data Visualization, Correlation plot, Scatter plot, MANOVA, Fisher Exact Test, P-value

Quantitative analysis of climate change and human activities on vegetation gross primary productivity in Nei Mongol, China

Quantitative analysis of climate change and human activities on vegetation gross primary productivity in Nei Mongol, China

A novel approach for enhanced abnormal action recognition via coarse and precise detection stage

<abstract> <p>With the proliferation of urban video surveillance systems, the abundance of surveillance video data has emerged as a pivotal asset for enhancing public safety. Within these video archives, the identification of abnormal human actions carries profound implications for security incidents. Nevertheless, existing surveillance systems primarily rely on conventional algorithms, leading to both missed incidents and false alarms. To address the challenge of automating multi-object surveillance video analysis, this study introduces a comprehensive method for the detection and recognition of multi-object abnormal actions. This study comprises a two-stage framework: the coarse detection stage employs an enhanced YOWOv2E model for spatio-temporal action detection, while the precise detection stage utilizes a two-stream network for precise action classification. In parallel, this paper presents the PSA-Dataset to address the current limitations in the field of abnormal action detection. Experimental results, collected from both public datasets and a self-built dataset, illustrate the effectiveness of the proposed method in identifying a wide spectrum of abnormal actions. This work offers valuable insights for automating the analysis of human actions in videos pertaining to public security.</p> </abstract>

Fundamentos do processo de trabalho em terapia ocupacional: uma abordagem analítica a partir do diálogo entre Terapia Ocupacional Social e Saúde Coletiva Latino-Americana

Abstract Introduction This paper presents an analytical approach to occupational therapy practice grounded in historical dialectical materialism. It understands occupational therapy professional practice as work within a capitalist system, and thus subject to analysis through the labor process theory. Objective To propose an analytical approach to occupational therapy practice aimed at supporting the development of an emancipatory practice framework based on the dialogue between Social Occupational Therapy and Latin American Collective Health. Method Utilizing emancipatory action research methodologies, this study conducted 10 workshops with 10 occupational therapists aimed at collective knowledge production. Results Human activity is identified as the object of occupational therapy working process, defined by the category work in its ontological dimension. Radical participation is set as the intended product of this process, which entails engaging in efforts to transform the social determination of collective living conditions. The individuals in this process include occupational therapists and the followed-up individuals, groups, and/or communities, all viewed as social beings collaborating toward social transformation. Four theoretical-methodological tools are identified: emancipatory analysis of human activities, emancipatory proposition of human activities, operationalization of radical participation, and interventions in the social fabric. Conclusion Occupational therapy, as a form of social practice, offers a specific contribution to the development of a collective revolutionary praxis. This praxis, aimed at realizing a concrete utopia, seeks to generate the means to create a new society.

DGU-HAO: A Dataset With Daily Life Objects for Comprehensive 3D Human Action Analysis

DGU-HAO: A Dataset With Daily Life Objects for Comprehensive 3D Human Action Analysis

U-Transformer-based multi-levels refinement for weakly supervised action segmentation

Action segmentation is a research hotspot in human action analysis, which aims to split videos into segments of different actions. Recent algorithms have achieved great success in modeling based on temporal convolution, but these methods weight local or global timing information through additional modules, ignoring the existing long-term and short-term information connections between actions. This paper proposes a U-Transformer structure based on multi-level refinement, introduces neighborhood attention to learn the neighborhood information of adjacent frames, and aggregates video frame features to effectively process long-term sequence information. Then a loss optimization strategy is proposed to smooth the original classification effect and generate a more accurate calibration sequence by introducing a pairing similarity optimization method based on deep feature learning. In addition, we propose a timestamp supervised training method to generate complete information for actions based on pseudo-label predictions for action boundary predictions. Experiments on three challenging action segmentation datasets, 50Salads, GTEA, and Breakfast, show that our model performs state-of-the-art models, and our weakly supervised model also performs comparably to fully supervised performance.

Suitability analysis of human activities over Antarctic ice shelves: an integrated assessment of natural conditions based on machine learning algorithms

ABSTRACT Human activities increase significantly over Antarctic ice shelves. However, they are constantly faced with danger posed by the harsh environment. For decision-making, it is a prerequisite to have the macro-scale suitability information about human activity site selection. Here, we define a new index, the human activity suitability (HAS) index, to quantitatively analyze the locational suitability of human activity sites over Antarctic ice shelves. Two multi-criteria decision analysis methods (AHP + Entropy, TOPSIS) and three machine learning methods (support vector machine, random forest and logistic regression) are tested to develop HAS maps. Nine conditioning factors about ice surface features, ice shelf stability, meteorology and topography are generated as input parameters. The accuracy of the proposed models is evaluated using metrics such as the area under curve (AUC-ROC), root mean square error, overall accuracy and kappa index. The results indicate that the Random Forest performs best. The HAS map exhibits great heterogeneity driven by the synergistic influence of multiple factors. Areas in low HAS classes are concentrated at Crosson, Brunt, Thwaites and the edge of some ice shelves, implying the complex environment in these regions. The findings can provide a new insight for forecasting the potential human footprint and support sustainability research in Antarctica.

DGU-HAU: A Dataset for 3D Human Action Analysis on Utterances

Constructing diverse and complex multi-modal datasets is crucial for advancing human action analysis research, providing ground truth annotations for training deep learning networks, and enabling the development of robust models across real-world scenarios. Generating natural and contextually appropriate nonverbal gestures is essential for enhancing immersive and effective human–computer interactions in various applications. These applications include video games, embodied virtual assistants, and conversations within a metaverse. However, existing speech-related human datasets are focused on style transfer, so they have limitations that make them unsuitable for 3D human action analysis studies, such as human action recognition and generation. Therefore, we introduce a novel multi-modal dataset, DGU-HAU, a dataset for 3D human action on utterances that commonly occurs during daily life. We validate the dataset using a human action generation model, Action2Motion (A2M), a state-of-the-art 3D human action generation model.

Connecting bones at Schöningen 13II-4 “Spear Horizon”: an analysis of site formation and human activity through faunal refitting

Connecting bones at Schöningen 13II-4 “Spear Horizon”: an analysis of site formation and human activity through faunal refitting

Study on groundwater pollution and its human impact analysis using geospatial techniques in semi-urban of south India

Groundwater recharging and thus renewable groundwater supplies will experience considerable changes due to climate change. The present investigation focussed to evaluate the susceptibility of subsurface water in rapid developing textile industrial region of Southern India. To determine the aquifer's susceptibility, the DRASTIC-LU Remote sensing and GIS based vulnerability model was applied and the results were compared with human activity risk (HAR) analysis to ensure the water borne disease on human health. DRASTIC layers like Depth to Water, Recharge, Aquifer Media, Soil Media, Topography, Impact of Vadose Zone, and Conductivity of the Aquifer are overlay along with respective LULC layers (2010 & 2020) to generate the DRASTIC-LU models for 2010 and 2020. The results of the DRASTIC LU groundwater vulnerability pattern between the range of 25–201 and 55 to 212 in 2010 and 2020, and have been classified into five categories from very high to low susceptibility. The effects of various Drastic-LU parameters were 69%, 76%, 81%, 82%, 63%, 79%, and 84%, respectively, based on the output of the sensitivity analysis method. The areas of Northern, Central, Central East and Central Western portions were identified as maximum and very highly vulnerable on the DRASTIC-LU index map and rest of the locations were classed as low and medium in terms of vulnerability. The outcomes of a risk analysis of human activity revealed that areas with high human activity density, such as of Northern, Central, Central East and Central Western portions were shown to be a vulnerable between the years 2010 and 2020, according to the results of human activity risk analysis.

A Comparative Study of Supervised and Unsupervised Approaches in Human Activity Analysis Based on Skeleton Data

A Comparative Study of Supervised and Unsupervised Approaches in Human Activity Analysis Based on Skeleton Data

Quantitative Analysis of Climate Variability and Human Activities on Vegetation Variations in the Qilian Mountain National Nature Reserve from 1986 to 2021

Rapid climate variability and intense human activities generate obvious impacts on the Qilian Mountains ecosystem. The time series of fractional vegetation coverage (FVC) from 1986 to 2021 were used to quantify the impact of climate variability and human activities on vegetation variations in the Qilian Mountain National Nature Reserve (QMNNR), using 3147 land satellite images based on the Google Earth Engine cloud platform. The contributions of climate variability and human activities to FVC were quantified using multiple regression residual analysis. Partial correlation and correlation methods were used to quantify the impact of temperature, precipitation, and human activity footprints on FVC. The results showed that from 1986 to 2021, the increase rate of FVC was 1.7 × 10−3 y−1, and the high vegetation coverage of the FVC was mainly distributed in the southeastern part of the reserve. In contrast, the low vegetation coverage was mainly distributed in the northwest part of the reserve. The Mann–Kendall mutation test found that the year of 2009 was the year of the mutation. The growth rate of FVC from 2010 to 2021 was greater than that from 1986 to 2009. In addition, climate variability and human activities exhibited a remarkable spatial heterogeneity in FVC changes. Climate variability and human activities contributed 49% and 51% to the increase in FVC in the reserve, respectively, and the contribution of human activities was greater than that of climate variability. The warming and humidification phenomena in the reserve were obvious. However, precipitation was the dominant factor affecting the dynamic changes in FVC. This study improves our understanding of the response of vegetation dynamics to the climate and human activities in the QMNNR.