Related Tasks Research Articles

Reference Twice: A Simple and Unified Baseline for Few-Shot Instance Segmentation.

Few-Shot Instance Segmentation (FSIS) requires detecting and segmenting novel classes with limited support examples. Existing methods based on Region Proposal Networks (RPNs) face two issues: 1) overfitting suppresses novel class objects and 2) dual-branch models require complex spatial correlation strategies to prevent spatial information loss when generating class prototypes. We introduce a unified framework, Reference Twice (RefT), to exploit the relationship between support and query features for FSIS and related tasks. Our three main contributions are: 1) a novel transformer-based baseline that avoids overfitting, offering a new direction for FSIS; 2) demonstrating that support object queries encode key factors after base training, allowing query features to be enhanced twice at both feature and query levels using simple cross-attention, thus avoiding complex spatial correlation interaction; and 3) introducing a class-enhanced base knowledge distillation loss to address the issue of DETR-like models struggling with incremental settings due to the input projection layer, enabling easy extension to incremental FSIS. Extensive experimental evaluations on the COCO dataset under three FSIS settings demonstrate that our method performs favorably against existing approaches across different shots, e.g., +8.2/+9.4 performance gain over state-of-the-art methods with 10/30-shots.

Classifiability analysis of spectroscopic profiling datasets in food safety related discriminative tasks.

Discriminative tasks, i.e., the identification of different food materials, brands, and origins, have become an essential part of food safety control. In recent years, spectroscopic profiling combined with machine learning is becoming popular for food-related discriminative tasks, but finding an appropriate classification model can be challenging. Compared to the current "trial-and-error" practice, this paper proposes a dedicated two-step classifiability analysis framework to address this issue. The first step collects more than 90 diversified metrics to measure the dataset separability from different perspectives. The second step synthesizes these metrics into a quantitative score using meta-learner and decomposition-based strategies. Finally, two Raman spectroscopic profiling case studies were conducted to validate the method, demonstrating higher scores for the easily separable liquor dataset (around 1.0) compared to the more challenging table salt dataset (< 0.5). This score can guide researchers to determine the required model complexity and assess the adequacy of the current physio-chemical profiling instrument. We expected the classifiability analysis framework proposed in this research to be generalized to a wide range of machine learning applications within the realm of food, where data-driven classification or discriminative tasks are involved.

Recent kernel methods for interacting particle systems: first numerical results

Abstract Interacting particle systems (IPSs) are a very important class of dynamical systems, arising in different domains like biology, physics, sociology and engineering. In many applications, these systems can be very large, making their simulation and control, as well as related numerical tasks, very challenging. Kernel methods, a powerful tool in machine learning, offer promising approaches for analyzing and managing IPS. This paper provides a comprehensive study of applying kernel methods to IPS, including the development of numerical schemes and the exploration of mean-field limits. We present novel applications and numerical experiments demonstrating the effectiveness of kernel methods for surrogate modelling and state-dependent feature learning in IPS. Our findings highlight the potential of these methods for advancing the study and control of large-scale IPS.

Acute Effects of High-Intensity Resistance Exercise on Recognition of Relational Memory, Lactate, and Serum and Plasma Brain-Derived Neurotrophic Factor.

Baumgartner, NW, Belbis, MD, Kargl, C, Holmes, MJ, Gavin, TP, Hirai, DM, and Kao, S-C. Acute effects of high-intensity resistance exercise on recognition of relational memory, lactate, and serum and plasma brain-derived neurotrophic factor. J Strength Cond Res 38(11): 1867-1878, 2024-Acute aerobic exercise improves memory, but this phenomenon is understudied in response to resistance exercise (RE) despite evidence that RE-induced increases in lactate and brain-derived neurotrophic factor (BDNF) play mechanistic roles in memory performance. To determine the acute effect of RE on lactate, BDNF, and their associations with object and relational memory, blood lactate, and serum and plasma BDNF were taken from 36 adults (average age 23.64 ± 3.89 years; 18 woman) before and immediately after 42 minutes of high-intensity RE and a rest condition on counterbalanced days. Subjects then immediately studied a series of paired objects and completed object and relational recognition tasks. Results revealed a condition by trial interaction, previously studied objects were remembered less accurately following RE ( d = 0.66) but recognition occurred faster ( d = 0.28), indicating a speed-accuracy tradeoff following RE. There was no effect of either intervention on relational recognition performance. Lactate ( d = 3.68) and serum BDNF ( d = 0.74) increased following RE, whereas there was no time-related change in lactate and serum BDNF following rest. However, changes in lactate and BDNF did not predict any measures of object ( rs < 0.25, p s > 0.16) or relation recognition ( rs < 0.28, p s > 0.13). Collectively, these findings suggest that acute high-intensity RE selectively improves the processing speed of recognizing objects at the cost of less accurate recognition of previously studied objects. Furthermore, changes in object and relational memory performance are unlikely driven by acute increases in lactate or BDNF following high-intensity RE.

WACARDIA: graphical MATLAB software for Wireless Assessment of CARDiac Interoceptive Accuracy.

Many theories of psychological function emphasize the importance of bodily sensations and the ability to accurately detect them, known as interoceptive accuracy. The most common measure of interoceptive accuracy uses heartbeat detection tasks such as the Whitehead Task, yet to our knowledge there are no freely accessible programs to conduct this task. In this paper, we present novel software called WACARDIA (Wireless Assessment of CARDiac Interoceptive Accuracy), which is free, open-source software that conducts the heartbeat detection task using Matlab and Psychtoolbox. WACARDIA contains several key features supporting participant engagement, operator convenience, and measurement accuracy. First, the program includes an optional practice trial of unlimited duration, a participant-facing graphical interface, and the ability to perform heartbeat detection training. Second, the operator is provided with a graphical user interface, live trial feedback, an accurate wireless electrocardiogram device, and a separate program to conduct the related Heartbeat Tracking task. Finally, the program ensures the accuracy of collected data by scheduling the delivery of tones with high precision and implementing fail-safes to automatically reset erroneous measurements. This paper includes flowcharts that help create transparency by describing our algorithm. We also outline customizable aspects of the program with the intent to have WACARDIA's algorithm expanded to accommodate more situations and applications. With this paper, we hope to encourage the practice of publicizing research software to contribute to the transparency, rigor, and reproducibility of scientific studies. WACARDIA and video tutorials are available at www.github.com/iankleckner/wacardia and http://wacardia.iankleckner.com.

Managing Overload: Tactics for Helping Leaders Combat Chronic Overwhelm

This article explores research-backed strategies for dealing with the chronic feelings of being overwhelmed that affect an estimated 75% of today's business professionals according to existing research. With overwhelm linked to negative consequences such as burnout, poor decision-making, and an inability to meet organizational goals, the article presents four key tactics drawn from research that leaders can use to better manage their workload and regain a sense of control. The strategies discussed include conducting an audit of current responsibilities to identify unnecessary tasks that can be cut, limited multitasking and interruptions through setting distraction-free periods allowing focus on priorities, empowering others through clear and appropriate delegation to reduce reliance on self, and instituting effective time management practices such as blocking out periods for priority work and batching related tasks. Case studies and findings are discussed for each approach. By adjusting perspectives and optimizing processes using evidence-based methods such as these, the article argues overwhelmed leaders can significantly reduce their chronic feelings of being overworked in support of both strong leadership performance and well-being within today's fast-paced business environment.

Combining transfer learning and statistical measures to predict performance of composite materials with limited data

AbstractPredicting the performance of composite materials is crucial for their application in civil infrastructure, yet limited experimental data often hinder the development of accurate and generalizable models. This study introduces a deep neural network (DNN) approach that combines summarizing statistics (SS) and transfer learning (TL)—termed the SSTL‐DNN approach—to address data scarcity in modeling composite materials. The computational novelty lies in the SS method's ability to extract comprehensive information from limited datasets by converting complex constitutive laws into concise statistical representations, thereby enabling efficient and effective model training. Simultaneously, the TL method enhances computational efficiency by leveraging knowledge from related tasks with abundant data to improve learning in the target task with scarce data. This combination not only reduces dependency on large datasets but also significantly improves model generalization. The proposed SSTL‐DNN approach is validated through two case studies: fiber‐reinforced polymer confined concrete and engineered cementitious composites. In both case studies, the SSTL‐DNN model reduces the required dataset size by up to 75% and decreases the validation error by 39%, compared to traditional deep learning models. These results demonstrate that the SSTL‐DNN approach not only overcomes data scarcity but also provides accurate predictions and generalization to unseen data, offering a practical solution for modeling composite materials with limited data.

Advanced Transfer Learning Strategies for Improving Deep Learning Model Efficiency and Accuracy

Machine learning primarily focuses on supervised learning, which relies on labeled data, while unsupervised learning addresses data without labels. Reinforcement learning, on the other hand, develops agents that learn optimal actions based on past experiences. Transfer learning, distinctively, enhances learning in a target task by leveraging knowledge from related source tasks. In recent years, many machine learning models have remained single-task focused. This paper explores scenarios in which transfer learning can be applied effectively from a source domain to a target domain, employing feature selection, extraction, and construction techniques. These methods enable transformation into a new feature space by generating a refined subset of the original features.

A multi-task spatio-temporal fusion network for offshore wind power ramp events forecasting

With the accelerated development of offshore wind farms, the impact of Wind Power Ramp Events (WPRE) on power systems has become more pronounced. Accurate prediction of WPRE is essential for mitigating their adverse effects on the grid. However, current studies on offshore WPRE are limited, especially regarding their spatio-temporal correlations and variability across large wind farm clusters. To address this, a novel forecasting approach using a multi-task spatio-temporal fusion network is presented. This method improves WPRE prediction by integrating spatial and temporal data and utilizing multitask learning to characterize ramp features. Firstly, an index system for WPRE characterization is developed, including metrics such as power change rate and duration. Based on this, an X-means classification method for WPRE is established. Secondly, a spatio-temporal encoder combining graph convolutional networks and temporal attention mechanisms has been proposed to model the spatiotemporal dependencies of meteorological changes in offshore wind farms. Finally, a multitask learning framework has been introduced to predict WPRE characteristic indices and occurrence probabilities. This approach enhances feature extraction efficiency through the joint feedback of multiple related tasks, thereby improving prediction accuracy. A case study using data from five offshore wind farms in Eastern China validates the efficacy of the proposed method.

Automatic Negation Detection for Semantic Analysis in Arabic Hotel Reviews Through Lexical and Structural Features: A Supervised Classification

One significant challenge in sentiment analysis is the presence of negation, which reverses the meanings of sentences, transformingpositive statements into negative ones and impacting the sentiment conveyed in the text. This issue is particularly pronounced in Arabic, a language known for its complex morphology. Detecting negation is crucial for enhancing sentiment analysis performance and various natural language processing applications. This paper presents an approach for automatically detecting negation in user-generated Arabic hotel reviews through lexical and structural features. It comprises several stages: data collection, text pre-processing, feature extraction, supervised learning classification, and evaluation. The study employed multiple supervised classification techniques, including naïve Bayes, random forest, logistic regression, support vector machines, and deep learning, to analyse lexical and structural features extracted from the dataset. The results of the experiments yielded promising outcomes, demonstrating the feasibility of the approach for practical applications. The classifiers exhibited highly comparable performance in identifying negation, with only marginal deviations in their performance metrics. Notably, the deep learning classifier consistently emerged as the top performer, achieving an exceptionally high overall accuracy rate of 99.24 percent, surpassing established benchmarks in Arabic text processing and underscoring its potential for practical applications. These findings hold significant implications for advancing Arabic text processing, particularly in sentiment analysis and related NLP tasks. The high accuracy of 99.24 percent achieved by the deep learning classifier highlights its robustness in accurately detecting negation, a critical challenge in sentiment analysis. This classifier performance demonstrates the potential to be integrated into real-world applications, such as automated review systems and opinion mining tools, where accurate sentiment interpretation is essential.

Modeling takeover time for non-driving related tasks using hybrid deep neural networks in conditionally automated vehicles

To enhance the safety of the transition from autonomous to manual driving control, one of the key factors is quantifying the safe takeover transition, namely the takeover time. Previous studies have focused on factors influencing takeover time and predicting takeover time using traditional methods, but there is a lack of research utilizing deep learning networks and driver physiological factors for takeover time prediction. Toward this end, we propose a hybrid deep learning network model dedicated to predicting drivers’ takeover time with a dataset obtained from a bench experiment involving 46 drivers. Firstly, to address the potential information loss commonly associated with traditional manual feature extraction techniques, a convolutional neural network (CNN) is developed to automatically extract physiological spatial representations from drivers. Additionally, a double-layer bidirectional long short-term memory network (DBiLSTM) is incorporated to model the physiological time representation dimension, enabling a more refined representation of physiological signals. Then, to tackle the complexities introduced by individual driver variations, a self-attention mechanism is enabling the model to automatically prioritize critical features, thereby enhancing the accuracy and reliability of its predictions. The results demonstrate that the proposed approach surpasses traditional machine learning network models in predictive accuracy for takeover time prediction, thereby exhibiting novelty in the field of vehicle takeover control research. This provides a more reliable technical foundation for anticipating the durations of vehicle takeovers.

AI chatbot: I want help, not the answer!

In recent years, the integration of artificial intelligence (AI) into educational settings has become a topic of great interest. As technology continues to evolve, educators and students are exploring how AI can enhance learning experiences. University of South-Eastern Norway (USN) and Kristiania University College (HK) are developing a web-based educational tool, DbPersist, where IT students can practice tasks within the database subject and receive automated feedback. The latest version now also offers students suggestions on how to utilize AI for learning while they are solving various database related tasks. In this study, we explore the students' experiences with getting automatically generated “prompts” for further use in their preferred AI tool, whenever they request assistance. Do they see the value in gaining experience in using prompts designed for learning? Do they believe it might help them using AI for learning purposes in other courses in their study? To answer these questions, we conducted surveys and interviews. Participants included bachelor students across various disciplines and multiple campuses at USN. The participants were all enrolled in database courses with learning goals involving the database query language SQL and ER modelling (drawing Entity-Relationship Diagrams). Early findings suggest that students appreciate the assistance provided by DbPersist’s AI prompts. They value the emphasis on AI-assisted learning and find the prompts useful for getting assistance in solving difficult tasks as opposed to getting complete answers. They also see the value in gaining experience in using AI chatbots using prompts designed for learning and understanding. As AI continues to shape education, tools like DbPersist offer exciting possibilities. DbPersist exemplifies how AI may be used to empower students, emphasizing learning and understanding alongside correctness.

Exploring the roles and influences of the modern general practice receptionist: a systematic review with narrative synthesis.

Modern general practice is characterised by increased demand and growing multi-disciplinarity, including ring-fenced funding for additional non-clinical roles. However, for practice receptionists training has remained unchanged for decades, yet primary care is under greater pressure than ever with receptionists becoming a growing focal point of abuse and unprecedented numbers leaving the role. To present the evidence of the range of tasks that receptionists continue to perform, describing their impact on primary care delivery and how the role might be better supported. We conducted a systematic review of evidence contained in the major medical databases (Medline/PubMed, CINAHL, ASSIA, Cochrane Library, and EMBASE) up until July 2023, including hand searches of the bibliographies of included studies. We identified 28 studies which we grouped into three themes: Service delivery, confirming the continuing role of receptionists in providing administrative support alongside clinical tasks of prioritising patients for consultations, facilitating repeat prescriptions, and communicating blood test results; Patient attitudes describing how patient's lacked trust in receptionists viewed as unqualified and unnecessarily obstructive; Finally, in considering Receptionist experience, the contrast between their confidence in performing administrative roles and the anxiety induced from the clinically related tasks was described, particularly the mounting pressure from patients to meet their preferences for clinician appointments. Though confident performing administrative tasks, receptionists described the uncertainty and anxiety when providing clinically oriented support or managing patients when their requests for appointments could not be met. More appropriate training or professionalisation might improve staff retainment.

Semantic memory navigation in HIV: Conceptual associations and word selection patterns

Objective: This proof-of-concept study aimed to characterize semantic memory profiles in individuals with human immunodeficiency virus (HIV) and mild neurocognitive impairment. Method: Using a semantic relatedness task, we explored conceptual association and word selection patterns in people living with HIV (PLWH; n = 50) relative to people living without HIV (n = 46). We also studied whether word selection patterns in the PLWH group were associated with working memory capacity, cognitive flexibility and inhibitory control. Results: While accuracy did not differ between groups, PLWH produced significantly longer responses than controls (r = .32), with fewer hypernyms (d = .47), more troponyms (r = .37), and words that were more frequent (r = .39) and had more phonological neighbors (r = .22). These patterns survived covariation with participants’ cognitive status. None of these patterns correlated with measures of working memory, cognitive flexibility, inhibitory control or viral load (all correlation coefficients < .36). Conclusions: Together, these results suggest that PLWH might use alternative word finding strategies during semantic memory navigation, irrespective of the severity of other cognitive symptoms. Such findings contribute to the characterization of cognitive deficits in HIV and to the search for novel markers of the condition.

Women’s health and psychological well-being in the return-to-work process after long-term sick leave for common mental disorders: women’s and first-line managers’ perspectives

Background and aimCommon mental disorders are common reasons for long-term sick leave, especially among women. Return to work is often complex and unsuccessful, why more knowledge is needed regarding women’s health and psychological well-being in the return-to-work process. Therefore, the aim was to describe women’s health and psychological well-being in the return-to-work process, from women’s and first-line managers’ perspectives.MethodsIndividual interviews were conducted with 17 women and 16 first-line managers. Qualitative content analysis was performed based on the content areas “Women’s health (i.e. overall well-being, both physical and psychological) throughout the whole RTW process” and “Women’s psychological well-being (happiness, meaning and a sense of being significant) at work after work resumption” Themes and categories were created.ResultsWomen and managers had similar descriptions, i.e. that women’s health and psychological well-being depend on the individual characteristics of women themselves, their private life, work and other stakeholders. However, women described relational work tasks (e.g. meeting patients) as beneficial for health, and highlighted small stressors in the work environment, which the managers did not. Having work that was compatible with private life, being in good health, having stimulating work tasks and strengthening relationships at work were important for the women’s psychological well-being.ConclusionsBased on women’s and first-line managers experiences, promotion of women’s health and psychological well-being during the return-to-work process requires individually adapted assessments and actions involving women’s entire life situation. First-line managers should know that relational work tasks (e.g., meeting patients) can be beneficial for women’s health as well as that minor stressor in the work environment can put their health at risk.

Relational Job Crafting and Task Performance: The Mediating Role of Work Engagement

This study investigates the mediating role of work engagement on the effect of relational job crafting on task performance among employees in the public sector of Ghana. Relational job crafting involves offers significant benefits to public sector employees by enhancing their performance through collaborative engagement. Thus, when employees actively adjust their social work environment, they can reduce (increase) work-related stress (sense of purpose) through fostering stronger relationships, which also improve well-being and performance. The research employed a closed-ended questionnaire and conducted a cross-sectional online survey. The study’s target population consists of 17,163 public sector employees in Cape Coast Metropolis, out of which 456 were randomly selected to complete the survey. The sample size was estimated using the Adam, (2020) approach for selecting samples in a quantitative study. Data collection took place from January to June 2023, and the analysis was conducted utilising the R version of partial least squares structural equation modelling (PLS-SEM). The study revealed an average yet statistically significant influence of relational job crafting on both work engagement and task performance. Thus, relational job crafting had a significant positive effect on work engagement (β = 0.502, p &lt; 0.01) and task performance (β = 0.293, p &lt; 0.01), with work engagement further affecting task performance significantly (β = 0.475, p &lt; 0.01). The significant positive relationship shows that the increase in both relational job crafting and work engagement among Ghanaian public sector employees further improves their performance in the various tasks assigned to them. Also, despite the direct relationship between work engagement and relational job crafting and task performance, it was further revealed that the latter further mediate the relationship among the former and task performance. Thus, it was found that work engagement serves as a complementary partial mediator in the relationship between relational job crafting and task performance (β = 0.238, SE = 0.034, 95% CI = 0.177, 0.309). These findings support the initial hypotheses established for the study. In terms of policy recommendations, this study emphasises the importance of fostering both work engagement and relational crafting to enhance task performance in the Ghanaian public sector. It suggests that employers, including government bodies, directors, and chief executive officers of public institutions, as well as policymakers, including labour unions, should implement measures across various sectors of public services to strengthen employee relations and increase engagement, ultimately enhancing task performance

Want a ride? Attitudes towards autonomous driving and behavior in autonomous vehicles

SummaryResearch conducted previously has focused on either attitudes toward or behaviors associated with autonomous driving. In this paper, we bridge these two dimensions by exploring how attitudes towards autonomous driving influence behavior in an autonomous car. We conducted a field experiment with twelve participants engaged in non-driving related tasks. Our findings indicate that attitudes towards autonomous driving do not affect participants’ driving interventions in vehicle control and eye glance behavior. Therefore, studies on autonomous driving technology lacking field tests might be unreliable for assessing the potential behaviors, attitudes, and acceptance of autonomous vehicles.

Better together: Automated app review analysis with deep multi-task learning

Context:User reviews of mobile apps provide an important communication channel between developers and users. Existing approaches to automated app review analysis mainly focus on one task (e.g., bug classification task, information extraction task, etc.) at a time, and are often constrained by the manually defined patterns and the ignorance of the correlations among the tasks. Recently, multi-task learning (MTL) has been successfully applied in many scenarios, with the potential to address the limitations associated with app review mining tasks. Objective:In this paper, we propose MABLE, a deep MTL-based and semantic-aware approach, to improve app review analysis by exploiting task correlations. Methods:MABLE jointly identifies the types of involved bugs reported in the review and extracts the fine-grained features where bugs might occur. It consists of three main phases: (1) data preparation phase, which prepares data to allow data sharing beyond single task learning; (2) model construction phase, which employs a BERT model as the shared representation layer to capture the semantic meanings of reviews, and task-specific layers to model two tasks in parallel; (3) model training phase, which enables eavesdropping by shared loss function between the two related tasks. Results:Evaluation results on six apps show that MABLE outperforms ten commonly-used and state-of-the-art baselines, with the precision of 79.76% and the recall of 79.24% for classifying bugs, and the precision of 79.83% and the recall of 80.33% for extracting problematic app features. The MTL mechanism improves the F-measure of two tasks by 3.80% and 4.63%, respectively. Conclusion:The proposed approach provides a novel and effective way to jointly learn two related review analysis tasks, and sheds light on exploring other review mining tasks.

서비스산업 육성을 위한 정책추진의 성과와 한계, 그리고 전략적 접근 방향 모색

This study aims to identify the achievements, limitations, and future challenges of service industry policies implemented between 2000 and 2024, and to explore effective policy directions for fostering the service industry. Since the 2000s, the government, which had primarily focused on manufacturing-centric industrial policies, has recognized the service industry as another key pillar for the continued growth of our economy and has consistently made policy efforts to support its development. As a result, while the transition to a service-based economy and the high value-added nature of the service industry have not progressed rapidly, some improvements have been made. However, our service industry is still largely driven by numerous small and low-value-added businesses, and there is a need to expand the industrial foundation that will drive innovation. In addition, with the changing industrial environment, such as advances in digital technology, it is also necessary to enhance the potential capabilities that will shape the future of the service industry. To effectively promote the transition to a service-based economy, two approaches can be considered. One is to identify the issues raised during the service industry's growth and develop solutions. Strategies could include establishing a policy support system, exploring policies that reflect the characteristics of the service industry, and fostering a shift in perception toward the industry. The other approach involves identifying the potential capabilities that will drive the future of the service industry and supporting the foundation to enhance these capabilities. Examples of related tasks include supporting networking among participants, enhancing capabilities needed for developing new business models, and creating a service provision environment to ensure these models take root in the market.

Optimized biomedical entity relation extraction method with data augmentation and classification using GPT-4 and Gemini.

Despite numerous research efforts by teams participating in the BioCreative VIII Track 01 employing various techniques to achieve the high accuracy of biomedical relation tasks, the overall performance in this area still has substantial room for improvement. Large language models bring a new opportunity to improve the performance of existing techniques in natural language processing tasks. This paper presents our improved method for relation extraction, which involves integrating two renowned large language models: Gemini and GPT-4. Our new approach utilizes GPT-4 to generate augmented data for training, followed by an ensemble learning technique to combine the outputs of diverse models to create a more precise prediction. We then employ a method using Gemini responses as input to fine-tune the BioNLP-PubMed-Bert classification model, which leads to improved performance as measured by precision, recall, and F1 scores on the same test dataset used in the challenge evaluation. Database URL: https://biocreative.bioinformatics.udel.edu/tasks/biocreative-viii/track-1/.