基于LDA主题模型的GitHub Actions工作流项目推荐算法

在CI/CD实践中,自动化已成为软件开发实践中的一种规范。GitHub引入GitHub Actions为软件维护者提供自动化的持续集成工作流方案,尽管其为开发者提供了诸多便利,GitHub社区也提供了许多第三方的GitHub Actions服务,但仍然只有极少的项目在使用。为了满足开发人员对工作流自动化的需求,减少非开发任务工作量,提出一种基于隐含狄利克雷分布(LDA)主题模型和Jensen-Shannon距离的GitHub Actions工作流项目推荐算法。通过对GitHub Actions存储库的README文件进行主题建模,将GitHub的事件描述文本和用户输入作为模型输入,为正在开发的代码仓库推荐GitHub Actions服务。将该推荐模型与标准的基于余弦相似度的方法进行比较的实验结果表明,该方法能有效改善开源软件仓库的推荐精度。

Quantitative distinction of the relative actions of climate change and human activities on vegetation evolution in the Yellow River Basin of China during 1981-2019

Under the combined influence of climate change and human activities,vegetation ecosystem has undergone profound changes.It can be seen that there are obvious differences in the evolution patterns and driving mechanisms of vegetation ecosystem in different historical periods.Therefore,it is urgent to identify and reveal the dominant factors and their contribution rates in the vegetation change cycle.Based on the data of climate elements(sunshine hours,precipitation and temperature),human activities(population intensity and GDP intensity)and other natural factors(altitude,slope and aspect),this study explored the spatial and temporal evolution patterns of vegetation NDVI in the Yellow River Basin of China from 1989 to 2019 through a residual method,a trend analysis,and a gravity center model,and quantitatively distinguished the relative actions of climate change and human activities on vegetation evolution based on Geodetector model.The results showed that the spatial distribution of vegetation NDVI in the Yellow River Basin showed a decreasing trend from southeast to northwest.During 1981-2019,the temporal variation of vegetation NDVI showed an overall increasing trend.The gravity centers of average vegetation NDVI during the study period was distributed in Zhenyuan County,Gansu Province,and the center moved northeastwards from 1981 to 2019.During 1981-2000 and 2001-2019,the proportion of vegetation restoration areas promoted by the combined action of climate change and human activities was the largest.During the study period(1981-2019),the dominant factors influencing vegetation NDVI shifted from natural factors to human activities.These results could provide decision support for the protection and restoration of vegetation ecosystem in the Yellow River Basin.

Workout Action Recognition in Video Streams Using an Attention Driven Residual DC-GRU Network

Regular exercise is a crucial aspect of daily life, as it enables individuals to stay physically active, lowers thelikelihood of developing illnesses, and enhances life expectancy. The recognition of workout actions in videostreams holds significant importance in computer vision research, as it aims to enhance exercise adherence, enableinstant recognition, advance fitness tracking technologies, and optimize fitness routines. However, existing actiondatasets often lack diversity and specificity for workout actions, hindering the development of accurate recognitionmodels. To address this gap, the Workout Action Video dataset (WAVd) has been introduced as a significantcontribution. WAVd comprises a diverse collection of labeled workout action videos, meticulously curated toencompass various exercises performed by numerous individuals in different settings. This research proposes aninnovative framework based on the Attention driven Residual Deep Convolutional-Gated Recurrent Unit (ResDCGRU)network for workout action recognition in video streams. Unlike image-based action recognition, videoscontain spatio-temporal information, making the task more complex and challenging. While substantial progresshas been made in this area, challenges persist in detecting subtle and complex actions, handling occlusions,and managing the computational demands of deep learning approaches. The proposed ResDC-GRU Attentionmodel demonstrated exceptional classification performance with 95.81% accuracy in classifying workout actionvideos and also outperformed various state-of-the-art models. The method also yielded 81.6%, 97.2%, 95.6%, and93.2% accuracy on established benchmark datasets, namely HMDB51, Youtube Actions, UCF50, and UCF101,respectively, showcasing its superiority and robustness in action recognition. The findings suggest practicalimplications in real-world scenarios where precise video action recognition is paramount, addressing the persistingchallenges in the field. TheWAVd dataset serves as a catalyst for the development ofmore robust and effective fitnesstracking systems and ultimately promotes healthier lifestyles through improved exercise monitoring and analysis.

Recognition of Human Actions through Speech or Voice Using Machine Learning Techniques

The development of artificial intelligence(AI)and smart home technologies has driven the need for speech recognition-based solutions.This demand stems from the quest for more intuitive and natural interaction between users and smart devices in their homes.Speech recognition allows users to control devices and perform everyday actions through spoken commands,eliminating the need for physical interfaces or touch screens and enabling specific tasks such as turning on or off the light,heating,or lowering the blinds.The purpose of this study is to develop a speech-based classification model for recognizing human actions in the smart home.It seeks to demonstrate the effectiveness and feasibility of using machine learning techniques in predicting categories,subcategories,and actions from sentences.A dataset labeled with relevant information about categories,subcategories,and actions related to human actions in the smart home is used.The methodology uses machine learning techniques implemented in Python,extracting features using CountVectorizer to convert sentences into numerical representations.The results show that the classification model is able to accurately predict categories,subcategories,and actions based on sentences,with 82.99%accuracy for category,76.19%accuracy for subcategory,and 90.28%accuracy for action.The study concludes that using machine learning techniques is effective for recognizing and classifying human actions in the smart home,supporting its feasibility in various scenarios and opening new possibilities for advanced natural language processing systems in the field of AI and smart homes.

Feature Fusion-Based Deep Learning Network to Recognize Table Tennis Actions

A system for classifying four basic table tennis strokes using wearable devices and deep learning networks is proposed in this study.The wearable device consisted of a six-axis sensor,Raspberry Pi 3,and a power bank.Multiple kernel sizes were used in convolutional neural network(CNN)to evaluate their performance for extracting features.Moreover,a multiscale CNN with two kernel sizes was used to perform feature fusion at different scales in a concatenated manner.The CNN achieved recognition of the four table tennis strokes.Experimental data were obtained from20 research participants who wore sensors on the back of their hands while performing the four table tennis strokes in a laboratory environment.The data were collected to verify the performance of the proposed models for wearable devices.Finally,the sensor and multi-scale CNN designed in this study achieved accuracy and F1 scores of 99.58%and 99.16%,respectively,for the four strokes.The accuracy for five-fold cross validation was 99.87%.This result also shows that the multi-scale convolutional neural network has better robustness after fivefold cross validation.

Challenges and Countermeasures for Integrating the Protection of Environmental Rights into Actions for Addressing Climate Change

Climate change,which is the result of human activities,has wide-ranging impact.It poses a serious threat to human rights.Environmental rights are where the protection of the ecological environment and the development of human rights intersect.In view of the close relationship between the actions for addressing climate change and environmental rights,China should integrate the protection of environmental rights into the actions for addressing climate change,so as to achieve simultaneous development of both.In the process of coping with climate change,the right to climate stability that mainly pursues a“harmless”environment and the right to a more livable climate that pursues a“beautiful eco-environment”are specific manifestations of environmental rights and should be the priority of protection efforts.However,there are still some obstacles to achieving the coordinated development of the efforts to address climate change and the protection of environmental rights because traditional rights protection methods mainly give individuals subjective rights with the power to claim and are thus difficult to meet the needs of environmental rights protection in the context of climate change,and there are inherent value differences between responding to climate change and the realization of other human rights.Building a multi-level national obligation system to address climate change,giving full play to the role of courts in responding to climate change through moderate judicial activism,and coordinating the efforts to cope with climate change and the development of human rights under the guidance of a holistic system view are effective ways to overcome the aforementioned difficulties.

Abnormal Action Recognition with Lightweight Pose Estimation Network in Electric Power Training Scene

Electric power training is essential for ensuring the safety and reliability of the system.In this study,we introduce a novel Abnormal Action Recognition(AAR)system that utilizes a Lightweight Pose Estimation Network(LPEN)to efficiently and effectively detect abnormal fall-down and trespass incidents in electric power training scenarios.The LPEN network,comprising three stages—MobileNet,Initial Stage,and Refinement Stage—is employed to swiftly extract image features,detect human key points,and refine them for accurate analysis.Subsequently,a Pose-aware Action Analysis Module(PAAM)captures the positional coordinates of human skeletal points in each frame.Finally,an Abnormal Action Inference Module(AAIM)evaluates whether abnormal fall-down or unauthorized trespass behavior is occurring.For fall-down recognition,three criteria—falling speed,main angles of skeletal points,and the person’s bounding box—are considered.To identify unauthorized trespass,emphasis is placed on the position of the ankles.Extensive experiments validate the effectiveness and efficiency of the proposed system in ensuring the safety and reliability of electric power training.

Unveiling health rights:A call to action for sex workers'HIV care in the Philippines

We are writing in response to the article titled“Addressing the needs and rights of sex workers for HIV healthcare services in the Philippines”[1].The article calls for attention on the significant challenges faced by sex workers in the Philippines in accessing HIV healthcare.We appreciate the article’s effort to examine these issues in depth.We would like to present a constant flow of thoughts in this letter while highlighting the positive aspects,potential obstacles,and additional points that contribute to this ongoing discussion.

Impact of Action Observation Therapy along with Usual Physiotherapy Intervention of Individual with Alzheimer’s Disease

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by cognitive impairments in the initial stage, which lead to severe cognitive dysfunction in the later stage. Action observation therapy (AOT) is a multisensory cognitive rehabilitation technique where the patient initially observes the actions and then tries to perform. The study aimed to examine the impact of AOT along with usual physiotherapy interventions to reduce depression, improve cognition and balance of a patient with AD. A 67 years old patient with AD was selected for this study because the patient has been suffering from depression, dementia, and physical dysfunction along with some other health conditions like diabetes and hypertension. Before starting intervention, a baseline assessment was done through the Beck Depression Inventory (BDI) tool, the Mini-Cog Scale, and the Berg Balance Scale (BBS). The patient received 12 sessions of AOT along with usual physiotherapy interventions thrice a week for four weeks, which included 45 minutes of each session. After four weeks of intervention, the patient demonstrated significant improvement in depression, cognition, and balance, whereas the BDI score declined from moderate 21/63 to mild 15/63 level of depression. The Mini-Cog score improved from 2/5 to 4/5, and the BBS score increased from 18/56 to 37/56. It is concluded that AOT along with usual physiotherapy intervention helps to reduce depression, improve cognition and balance of people with AD.

Securing Stock Transactions Using Blockchain Technology: Architecture for Identifying and Reducing Vulnerabilities Linked to the Web Applications Used (MAHV-BC)

This paper deals with the security of stock market transactions within financial markets, particularly that of the West African Economic and Monetary Union (UEMOA). The confidentiality and integrity of sensitive data in the stock market being crucial, the implementation of robust systems which guarantee trust between the different actors is essential. We therefore proposed, after analyzing the limits of several security approaches in the literature, an architecture based on blockchain technology making it possible to both identify and reduce the vulnerabilities linked to the design, implementation work or the use of web applications used for transactions. Our proposal makes it possible, thanks to two-factor authentication via the Blockchain, to strengthen the security of investors’ accounts and the automated recording of transactions in the Blockchain while guaranteeing the integrity of stock market operations. It also provides an application vulnerability report. To validate our approach, we compared our results to those of three other security tools, at the level of different metrics. Our approach achieved the best performance in each case.

BCCLR:A Skeleton-Based Action Recognition with Graph Convolutional Network Combining Behavior Dependence and Context Clues

In recent years,skeleton-based action recognition has made great achievements in Computer Vision.A graph convolutional network(GCN)is effective for action recognition,modelling the human skeleton as a spatio-temporal graph.Most GCNs define the graph topology by physical relations of the human joints.However,this predefined graph ignores the spatial relationship between non-adjacent joint pairs in special actions and the behavior dependence between joint pairs,resulting in a low recognition rate for specific actions with implicit correlation between joint pairs.In addition,existing methods ignore the trend correlation between adjacent frames within an action and context clues,leading to erroneous action recognition with similar poses.Therefore,this study proposes a learnable GCN based on behavior dependence,which considers implicit joint correlation by constructing a dynamic learnable graph with extraction of specific behavior dependence of joint pairs.By using the weight relationship between the joint pairs,an adaptive model is constructed.It also designs a self-attention module to obtain their inter-frame topological relationship for exploring the context of actions.Combining the shared topology and the multi-head self-attention map,the module obtains the context-based clue topology to update the dynamic graph convolution,achieving accurate recognition of different actions with similar poses.Detailed experiments on public datasets demonstrate that the proposed method achieves better results and realizes higher quality representation of actions under various evaluation protocols compared to state-of-the-art methods.

Abnormal Action Detection Based on Parameter-Efficient Transfer Learning in Laboratory Scenarios

Laboratory safety is a critical area of broad societal concern,particularly in the detection of abnormal actions.To enhance the efficiency and accuracy of detecting such actions,this paper introduces a novel method called TubeRAPT(Tubelet Transformer based onAdapter and Prefix TrainingModule).Thismethod primarily comprises three key components:the TubeR network,an adaptive clustering attention mechanism,and a prefix training module.These components work in synergy to address the challenge of knowledge preservation in models pretrained on large datasets while maintaining training efficiency.The TubeR network serves as the backbone for spatio-temporal feature extraction,while the adaptive clustering attention mechanism refines the focus on relevant information.The prefix training module facilitates efficient fine-tuning and knowledge transfer.Experimental results demonstrate the effectiveness of TubeRAPT,achieving a 68.44%mean Average Precision(mAP)on the CLA(Crazy LabActivity)small-scale dataset,marking a significant improvement of 1.53%over the previous TubeR method.This research not only showcases the potential applications of TubeRAPT in the field of abnormal action detection but also offers innovative ideas and technical support for the future development of laboratory safety monitoring technologies.The proposed method has implications for improving safety management systems in various laboratory environments,potentially reducing accidents and enhancing overall workplace safety.

Ethical and strategic challenges of AI weapons: A call for global action

1 Impending prophecy About a decade ago,theoretical physicist Stephen Hawking expressed extreme concern about the development of artificial intelligence(AI),stating that"the development of full artificial intelligence could spell the end of the human race.Concurrently,in 2014,entrepreneur Elon Musk expressed similar cautionary sentiments,suggesting that AI might surpass nuclear weapons in terms of danger.Despite these warnings,the prevailing sentiment at the time was largely skeptical.However,only a few years later,on May 30,2023,hundreds of artificial intelligence experts and other notable figures issued a 22-word statement[1]warning against the“risk of extinction.”

General Optimal Trajectory Planning:Enabling Autonomous Vehicles with the Principle of Least Action

This study presents a general optimal trajectory planning(GOTP)framework for autonomous vehicles(AVs)that can effectively avoid obstacles and guide AVs to complete driving tasks safely and efficiently.Firstly,we employ the fifth-order Bezier curve to generate and smooth the reference path along the road centerline.Cartesian coordinates are then transformed to achieve the curvature continuity of the generated curve.Considering the road constraints and vehicle dynamics,limited polynomial candidate trajectories are generated and smoothed in a curvilinear coordinate system.Furthermore,in selecting the optimal trajectory,we develop a unified and auto-tune objective function based on the principle of least action by employing AVs to simulate drivers’behavior and summarizing their manipulation characteristics of“seeking benefits and avoiding losses.”Finally,by integrating the idea of receding-horizon optimization,the proposed framework is achieved by considering dynamic multi-performance objectives and selecting trajectories that satisfy feasibility,optimality,and adaptability.Extensive simulations and experiments are performed,and the results demonstrate the framework’s feasibility and effectiveness,which avoids both dynamic and static obstacles and applies to various scenarios with multi-source interactive traffic participants.Moreover,we prove that the proposed method can guarantee real-time planning and safety requirements compared to drivers’manipulation.

Offline Reinforcement Learning with Constrained Hybrid Action Implicit Representation Towards Wargaming Decision-Making

Reinforcement Learning(RL)has emerged as a promising data-driven solution for wargaming decision-making.However,two domain challenges still exist:(1)dealing with discrete-continuous hybrid wargaming control and(2)accelerating RL deployment with rich offline data.Existing RL methods fail to handle these two issues simultaneously,thereby we propose a novel offline RL method targeting hybrid action space.A new constrained action representation technique is developed to build a bidirectional mapping between the original hybrid action space and a latent space in a semantically consistent way.This allows learning a continuous latent policy with offline RL with better exploration feasibility and scalability and reconstructing it back to a needed hybrid policy.Critically,a novel offline RL optimization objective with adaptively adjusted constraints is designed to balance the alleviation and generalization of out-of-distribution actions.Our method demonstrates superior performance and generality across different tasks,particularly in typical realistic wargaming scenarios.

HgaNets:Fusion of Visual Data and Skeletal Heatmap for Human Gesture Action Recognition

Recognition of human gesture actions is a challenging issue due to the complex patterns in both visual andskeletal features. Existing gesture action recognition (GAR) methods typically analyze visual and skeletal data,failing to meet the demands of various scenarios. Furthermore, multi-modal approaches lack the versatility toefficiently process both uniformand disparate input patterns.Thus, in this paper, an attention-enhanced pseudo-3Dresidual model is proposed to address the GAR problem, called HgaNets. This model comprises two independentcomponents designed formodeling visual RGB (red, green and blue) images and 3Dskeletal heatmaps, respectively.More specifically, each component consists of two main parts: 1) a multi-dimensional attention module forcapturing important spatial, temporal and feature information in human gestures;2) a spatiotemporal convolutionmodule that utilizes pseudo-3D residual convolution to characterize spatiotemporal features of gestures. Then,the output weights of the two components are fused to generate the recognition results. Finally, we conductedexperiments on four datasets to assess the efficiency of the proposed model. The results show that the accuracy onfour datasets reaches 85.40%, 91.91%, 94.70%, and 95.30%, respectively, as well as the inference time is 0.54 s andthe parameters is 2.74M. These findings highlight that the proposed model outperforms other existing approachesin terms of recognition accuracy.

Adjustment mechanism of blasting dynamic-static action in the water decoupling charge

Water decoupling charge blasting excels in rock breaking,relying on its uniform pressure transmission and low energy dissipation.The water decoupling coefficients can adjust the contributions of the stress wave and quasi-static pressure.However,the quantitative relationship between the two contributions is unclear,and it is difficult to provide reasonable theoretical support for the design of water decoupling blasting.In this study,a theoretical model of blasting fracturing partitioning is established.The mechanical mechanism and determination method of the optimal decoupling coefficient are obtained.The reliability is verified through model experiments and a field test.The results show that with the increasing of decoupling coefficient,the rock breaking ability of blasting dynamic action decreases,while quasi-static action increases and then decreases.The ability of quasi-static action to wedge into cracks changes due to the spatial adjustment of the blast hole and crushed zone.The quasi-static action plays a leading role in the fracturing range,determining an optimal decoupling coefficient.The optimal water decoupling coefficient is not a fixed value,which can be obtained by the proposed theoretical model.Compared with the theoretical results,the maximum error in the model experiment results is 8.03%,and the error in the field test result is 3.04%.

Developments in the study of Chinese herbal medicine's assessment index and action mechanism for diabetes mellitus

In traditional Chinese medicine(TCM),based on various pathogenic symptoms and the‘golden chamber’medical text,Huangdi Neijing,diabetes mellitus falls under the category‘collateral disease’.TCM,with its wealth of experience,has been treating diabetes for over two millennia.Different antidiabetic Chinese herbal medicines re-duce blood sugar,with their effective ingredients exerting unique advantages.As well as a glucose lowering effect,TCM also regulates bodily functions to prevent diabetes associated complications,with reduced side effects compared to western synthetic drugs.Chinese herbal medicine is usually composed of polysaccharides,saponins,al-kaloids,flavonoids,and terpenoids.These active ingredients reduce blood sugar via various mechanism of actions that include boosting endogenous insulin secretion,enhancing insulin sensitivity and adjusting key enzyme activity and scavenging free radicals.These actions regulate glycolipid metabolism in the body,eventually achiev-ing the goal of normalizing blood glucose.Using different animal models,a number of molecular markers are available for the detection of diabetes induction and the molecular pathology of the disease is becoming clearer.Nonetheless,there is a dearth of scientific data about the pharmacology,dose-effect relationship,and structure-activity relationship of TCM and its constituents.Further research into the efficacy,toxicity and mode of action of TCM,using different metabolic and molecular markers,is key to developing novel TCM antidiabetic formulations.

Experimental study on the movement of oil spill under freeze-thaw action

Oil leakages cause environmental pollution,economic losses,and even engineering safety accidents.In cold regions,researchers urgently investigate the movement of oil spill in soils exposed to freeze-thaw cycles.In this study,a series of laboratory model experiments were carried out on the migration of oil leakage under freeze-thaw action,and the distributions of the soil temperature,unfrozen water content,and displacement were analyzed.The results showed that under freeze-thaw action,liquid water in soils migrated to the freezing front and accumulated.After the pipe cracked,oil pollutants first gathered at one side of the leak hole,and then moved around.The pipe wall temperature affected the soil temperature field,and the thermal influence range below and transverse the pipe wall(35–40 cm)was larger than that above the pipe wall(8 cm)owing to the soil surface temperature.The leaked oil's temperature would make the temperature of the surrounding soil rise.Oil would inhibit the cooling of the soils.Besides,oil migration was significantly affected by the gravity and water flow patterns.The freeze-thaw action would affect the migration of the oil,which was mainly manifested as inhibiting the diffusion and movement of oil when soils were frozen.Unfrozen water transport caused by freeze-thaw cycles would also inhibit oil migration.The research results would provide a scientific reference for understanding the relationship between the movement of oil pollutants,water,and soil temperature,and for establishing a waterheat-mass transport model in frozen soils.

Research Progress on Purification Process, Content Determination and Pharmacological Action of Atractylodin

Atractylodis Rhizoma comes from the dry rhizome of Atractylis lancea or Atractylodes chinensis in the Compositae family,and it is suitable for preventing and treating diseases such as cold,edema,night blindness and rheumatic arthralgia.Atractylodin is the main active component extracted and isolated from Atractylodis Rhizoma.A large number of studies have found that atractylodin has excellent drug activity in improving gastrointestinal emptying,anti-inflammation,inhibiting malignant tumor and reducing blood lipid.In this paper,the purification process and pharmacological activity of Atractylodin were summarized to provide a theoretical basis for basic research,clinical application and further development and utilization of atractylodin.