Classical Training Research Articles

Применение технологий виртуальной реальности для подготовки пожарных и спасателей

In firefighting and rescue operations, practical training plays a primary role, allowing rescuers to be constantly ready to act in the event of major incidents, accidents or emergencies. In addition to classical training methods, namely training in an unsuitable environment for breathing. Virtual reality is increasingly being used in training in many fields of activity, given the fact that in a virtual environment it is easy to simulate dangers that cannot be created in reality (explosions, fires, falls from a great height), and thus overcome the fear of such dangers. The capabilities of virtual reality simulators allow you to create training scenarios that emphasize the degree of physical fitness of rescuers, taking into account their physiological parameters, which are monitored throughout the training.

A systematic review and network meta-analysis on the effectiveness of exercise-based interventions for reducing the injury incidence in youth team-sport players. Part 1: an analysis by classical training components

Objectives The primary purposes were (a) to estimate the pooled effects of injury prevention programs (IPPs) on reducing overall and some specific body regions (lower extremity, thigh, knee, and ankle) injury incidence rates (IIRs) and (b) to compare the effects of single- and multi-component IPPs on mitigating injury risk in youth team sport athletes. A secondary objective was to explore the individual effects of different components on these IIRs. Materials and methods Searches were performed up to 15 January 2024 in PubMed, Web of Science, SPORTDiscus, and Cochrane Library. Eligible criteria were: exercise-based interventions evaluated against a control group, overall IIRs were reported, and youth (≤19 years old) team sport players. Two reviewers extracted data and assessed trial quality using the Consolidated Standards of Reporting Trials (CONSORT) statement, the Physiotherapy Evidence Database Scale (PEDro), and a risk of bias tool (Cochrane Back and Neck Group). Pooled effects were calculated by Frequentist random effects pairwise and network meta-analyses. Results Twenty-one studies were included. IPPs reduced overall, lower extremities, thigh, knee, and ankle IIRs by an average of approximately 35%. Most of the IPPs demonstrated statistically significant risk mitigation effects for overall and lower extremity injuries compared to control group. Interventions comprised exclusively of strength ([IRR = 0.3 [95%CI = 0.10–0.93]) and flexibility (IRR = 0.49 [95%CI = 0.36–0.68]), as well as those including stability exercises, were the most effective measures for reducing injuries in youth team sports. Conclusions The implementation of current IPPs in training sessions for several weeks has shown to be an effective strategy for reducing the risk of injury in youth team sport athletes by one-third. Indirect evidence suggests that strength, flexibility, and stability might be exercise components with the highest risk mitigation effects; however, more research is crucial to confirm our estimates with direct evidence.

Effects of 4 weeks of device-induced normobaric intermittent hypoxia/hyperoxia training on the performance of elite cyclists: A pilot study

Introduction & Purpose Classic altitude training and other forms of hypoxia are regularly used in endurance training (Mujika et al., 2019). Due to corona exit restrictions in 2019-2020, device-induced normobaric hypoxia/hyperoxia was suggested as an alternative to classic altitude training. Intermittent hypoxia/hyperoxia training (IHHT) has been studied and used in various forms since the 1940s (Serebrovskaya, 2002) with the advantage to be used at the athlete’s training site (Burtscher, 2005). The aim of this pilot study was to test the effects of a IHHT intervention on aerobic performance, hematological parameters and hypoxia-induced factor-1α (HIF-1α). Method A total of 11 athletes (2 women) from the U23 and elite national squads of the Austrian Cycling Federation completed 16 IHHT units in addition to their normal training program over a period of 4 weeks as part of a randomized cross over intervention study (control – IHHT, IHHT – control). The IHHT units were carried out on a Cellgym CellAir® Gecko Plus device at rest without exercise following a predetermined protocol (Sessions 1-8: 5 min of hypoxia (12.0% oxygen) followed by 3 min of hyperoxia (34.0% oxygen) repeated five times. Sessions 9-16): 6 min of hypoxia (11.0% oxygen) followed by 3 min of hyperoxia (34.0% oxygen). During the 4 weeks of control period, participants should maintain their training schedule. Aerobic performance was determined before and after the intervention by spiroergometry. As secondary outcomes, changes in hematological parameters and protein levels of HIF-1α were determined from venous blood which was processed immediately after collection. Serum was stored at -80 °C until HIF-1 α determination by ELISA technique. Results Eleven out of 13 athletes took part in the pre – post IHHT measurements, but only three of them managed to participate in the measurements planned for the control period, making it impossible to compare the two conditions. Replacement was not possible due to the availability of candidates and the ongoing preparation for the season. After 4 weeks of intervention with 16 IHHT training sessions, no significant changes were observed in performance parameters such as VO2peak, relative VO2max, maximum power (watts) or threshold values (p > 0.05). There were also no significant changes in blood enzymes, lipid and iron metabolism (p > 0.05). However, a tendency towards increased values of haemoglobin (z = 1.738; p = 0.082) and a significant increase of erythrocyte counts (z = 2.193; p = 0.028) were detected. Other haematological parameters and HIF-1α did not change significantly (p > 0.05). Discussion & Conclusion In summary, four weeks of device-induced normobaric IHHT had a modest effect on haematological parameters, but with no detectable transfer to maximal or submaximal endurance performance in highly trained athletes. Although various studies have shown that the difference between normobaric and hypobaric hypoxia exposure is not limited to the partial pressure of oxygen, but Mounier and Brugniaux state that the physiological responses are in fact equivalent (Mounier & Brugniaux, 2012). However, due to the limitations of the competitive setting and the low sample size, the final interpretation of the study remains difficult. References Burtscher, M. (2005). Intermittierende Hypoxie: Höhenvorbereitung, Training, Therapie [Intermittent hypoxia: altitude preparation, training, therapy]. Schweizerische Zeitschrift für Sportmedizin und Sporttraumatologie, 53(2), 61-67. Serebrovskaya, T. V. (2002). Intermittent hypoxia research in the former Soviet Union and the commonwealth of independent states: History and review of the concept and selected applications. High Altitude Medicine & Biology, 3(2), 205-221. https://doi.org/10.1089/15270290260131939 Mounier, R., & Brugniaux, J. V. (2012). Counterpoint: Hypobaric hypoxia does not induce different responses from normobaric hypoxia. Journal of Applied Physiology, 112(10), 1784-1786. https://doi.org/10.1152/japplphysiol.00067.2012a Mujika, I., Sharma, A. P., & Stellingwerff, T. (2019). Contemporary periodization of altitude training for elite endurance athletes: A narrative review. Sports Medicine, 49, 1651-1669. https://doi.org/10.1007/s40279-019-01165-y

A Conversation with Michael Maliakel

Abstract: In a video interview with Journal of Singing Associate Editor Melissa Treinkman on March 29, 2024, Michael Maliakel discussed his career and his experience playing the title role in the Broadway show Aladdin . A first place winner of the 2014 NATS National Musical Theatre Competition, Michael talked about the impact of his classical voice training on his Broadway career, the constantly ascending tessitura of musical theatre roles, and why he believes representation matters. A QR code is included at the end of the article which contains the link to the recorded interview with Michael Maliakel, posted on the NATS YouTube channel.

A cardiac rehabilitation programme based on neuromuscular training improves the functional capacity of patients with acute coronary syndrome: a preliminary randomised controlled trial

ObjectivesTo evaluate the effects of a cardiac rehabilitation programme based on neuromuscular training (NMT) compared with classic rehabilitation strength training (CRST) in patients diagnosed with acute coronary syndrome (ACS). DesignPreliminary randomised, double-blinded, parallel clinical trial. SettingUniversity health clinic in Valencia, Spain. ParticipantsThirty patients with ACS, InterventionsPatients were assigned to one of two groups at random: the NMT group (n=15) and the CRST group (n=15). All patients attended the 20 sessions of the exercise programme. Main outcome measuresThe primary outcome was the Incremental Shuttle Walk Test (ISWT). The secondary outcomes were the Chester Step Test (CST), the 30-Second Chair Stand Test (30CST), and hip flexor dynamometry. Assessments were made at baseline, immediately post-treatment, and at 6-month follow-up. ResultsThe NMT group showed a greater improvement in the ISWT than the CRST group, both at post-treatment {mean 648 [standard deviation (SD) 197] vs 493 (SD 219), mean difference 155, 95% confidence interval (CI) -1 to 310} and at follow-up [732 (SD 183) vs 518 (SD 222), mean difference 214, 95% CI 61 to 367]. The secondary outcomes showed significant between-group differences in favour of the NMT group at 6-month follow-up, except for the 30CST. ConclusionsThese preliminary findings indicate that a cardiac rehabilitation programme based on NMT may improve functional capacity in terms of patient performance, cardiorespiratory fitness and muscle strength. Moreover, the improvements were maintained at medium-term follow-up. This could help improve the design of rehabilitation sessions, considering factors associated with performing everyday activities in patients affected by ACS.

Is there room for cultural expression in classical dance training?

Is there room for cultural expression in classical dance training?

"Effect of Progressive Structured Neuromuscular Training Program on Balance, Performance and Injury Risk Prediction in Amateur Futsal Team Players: Protocol for a Randomized Controlled Trial"

"Neuromuscular training improves neuromuscular control and balance, consequently enhancing joint stability. This study aims to assess the impact of a structured neuromuscular training program on balance and performance among university amateur male futsal team players and to analyze biomechanical factors pertinent to injury risk assessments. Thirty athletes aged between 18 and 25, participating in the university male futsal team, will be enrolled in this study. At the outset, all athletes will undergo evaluation using the Korebalance® System, Countermovement Jump, Running-Based Anaerobic Sprint Test, Shot on Goal Test, Sit and Reach Test, Shuttle, Y Balance-Lower Quadrant, and The Tuck Jump test. Subsequently, they will maintain their regular training regimen of classical training twice a week for eight weeks and undergo reassessment using the same battery of tests after 8 weeks of period. Following this initial phase, the athletes will be randomly assigned to two groups: Control Group (n = 15) and Neuromuscular Training Group (n = 15). The Control Group (CG) will continue with their classical futsal training twice a week for eight weeks, while the Neuromuscular Training Group (NMEG) will engage in progressive structured neuromuscular training in addition to their regular futsal training, also twice a week for eight weeks. At the conclusion of the eight-week intervention period, both groups will undergo reevaluation using the same assessment methods. The study commenced in October 2023 and is anticipated to conclude by August 2024. It is anticipated that the integration of progressive structured neuromuscular training into futsal team training programs will yield positive outcomes."

Acute performance fatigability following continuous versus intermittent cycling protocols is not proportional to total work done.

Classical training theory postulates that performance fatigability following a training session should be proportional to the total work done (TWD); however, this notion has been questioned. This study investigated indices of performance and perceived fatigability after primary sessions of high-intensity interval training (HIIT) and constant work rate (CWR) cycling, each followed by a cycling time-to-task failure (TTF) bout. On separate days, 16 participants completed an incremental cycling test, and, in a randomized order, (i) a TTF trial at 80% of peak power output (PPO), (ii) an HIIT session, and (iii) a CWR session, both of which were immediately followed by a TTF trial at 80% PPO. Central and peripheral aspects of performance fatigability were measured using interpolated twitch technique, and perceptual measures were assessed prior to and following the HIIT and CWR trials, and again following the TTF trial. Despite TWD being less following HIIT (P=0.029), subsequent TTF trial was an average of 125s shorter following HIIT versus CWR (P<0.001), and this was accompanied by greater impairments in voluntary and electrically evoked forces (P<0.001), as well as exacerbated perceptual measures (P<0.001); however, there were no differences in any fatigue measure following the TTF trial (P≥0.149). There were strong correlations between the decline in TTF and indices of peripheral (r=0.70) and perceived fatigability (r≥0.80) measured at the end of HIIT and CWR. These results underscore the dissociation between TWD and performance fatigability and highlight the importance of peripheral components of fatigability in limiting endurance performance during high-intensity cycling exercise.

Electrotactile BCI for Top-Down Somatosensory Training: Clinical Feasibility Trial of Online BCI Control in Subacute Stroke Patients.

This study investigates the feasibility of a novel brain-computer interface (BCI) device designed for sensory training following stroke. The BCI system administers electrotactile stimuli to the user's forearm, mirroring classical sensory training interventions. Concurrently, selective attention tasks are employed to modulate electrophysiological brain responses (somatosensory event-related potentials-sERPs), reflecting cortical excitability in related sensorimotor areas. The BCI identifies attention-induced changes in the brain's reactions to stimulation in an online manner. The study protocol assesses the feasibility of online binary classification of selective attention focus in ten subacute stroke patients. Each experimental session includes a BCI training phase for data collection and classifier training, followed by a BCI test phase to evaluate online classification of selective tactile attention based on sERP. During online classification tests, patients complete 20 repetitions of selective attention tasks with feedback on attention focus recognition. Using a single electroencephalographic channel, attention classification accuracy ranges from 70% to 100% across all patients. The significance of this novel BCI paradigm lies in its ability to quantitatively measure selective tactile attention resources throughout the therapy session, introducing a top-down approach to classical sensory training interventions based on repeated neuromuscular electrical stimulation.

Maximizing Participation in Olfactory Training in a Sample with Post-COVID-19 Olfactory Loss.

This study aims to highlight the feasibility of an olfactory training program entirely monitored through online media in COVID-19 patients. Classic olfactory training was performed with a sample with olfactory loss due COVID-19 (n = 11). Participants were engaged on a weekly video call in order to improve adherence and collect information regarding the number of correct answers and the individuals' perception of olfactory function. The olfactory status after training was compared to two groups, one composed of participants who contracted COVID-19 but did not report olfactory loss (n = 11) and a sample composed of healthy participants (n = 11). The experimental group showed improvements throughout the training period (TDI score on week 0 was 20.3 (5.6) and 24.6 (4.3) for week 12, and on week 24 was 25.4 (6.2) (F = 5.115, df = 2, 20, p = 0.016), and post hoc tests showed that participants significantly improved their TDI score in W12 compared to W0 (SMD = 0.869, p = 0.041) and in W24 compared to W0 (SMD = 0.859, p = 0.041). The experimental group showed lower scores when compared with both groups, and the no OT COVID-19 group showed lower scores than the healthy control group, even though they did not report olfactory alterations. Findings suggest that the strategies applied to improve adherence were successful since 100% of the sample completed the training adherence, offering a valuable framework for future olfactory training studies.

BRIEF LITERATURE REVIEW ON RUMBA MOVEMENT ANALYSIS

The analysis of sports dance mechanics is a topic of interest among researchers in recent years. Pengping Zhu (2004) studied classical ballet training combined with biodynamics and analyzed its correlation. Xiling Deng (2001) discussed mechanical issues in aerobic gymnastics. Latin dancers have radically different biomechanical demands than standard dancers (McCabe, T. R., Wyon, M., Ambegaonkar, J. P., Redding, E., 2013). In Latin American dances, both men and women follow a similar step pattern and often mirror each other's movements (Laird, W. &amp; Laird, J., 2003). In order to be able to understand the motor content that supports the technical executions specific to sports dance, we need to be guided by the movement analysis models. In this sense Bota A. (2007), citing Rasch and Burke, presents stages of the model of biomechanical analysis of movement. Due to the complexity of dance steps and specific combinations, sports dance is an extremely demanding field, being approached by the performance field, so it requires many hours of training to achieve general and specific physical training.

Can Multisensory Olfactory Training Improve Olfactory Dysfunction Caused by COVID-19?

Approximately 30-60% of people suffer from olfactory dysfunction (OD) such as hyposmia or anosmia after being diagnosed with COVID-19; 15-20% of these cases last beyond resolution of the acute phase. Previous studies have shown that olfactory training can be beneficial for patients affected by OD caused by viral infections of the upper respiratory tract. The aim of the study is to evaluate whether a multisensory olfactory training involving simultaneously tasting and seeing congruent stimuli is more effective than the classical olfactory training. We recruited 68 participants with persistent OD for two months or more after COVID-19 infection; they were divided into three groups. One group received olfactory training which involved smelling four odorants (strawberry, cheese, coffee, lemon; classical olfactory training). The other group received the same olfactory stimuli but presented retronasally (i.e., as droplets on their tongue); while simultaneous and congruent gustatory (i.e., sweet, salty, bitter, sour) and visual (corresponding images) stimuli were presented (multisensory olfactory training). The third group received odorless propylene glycol in four bottles (control group). Training was carried out twice daily for 12 weeks. We assessed olfactory function and olfactory specific quality of life before and after the intervention. Both intervention groups showed a similar significant improvement of olfactory function, although there was no difference in the assessment of quality of life. Both multisensory and classical training can be beneficial for OD following a viral infection; however, only the classical olfactory training paradigm leads to an improvement that was significantly stronger than the control group.

Characterization of Gut Microbiota in Rats and Rhesus Monkeys After Methamphetamine Self-administration.

Methamphetamine (MA) is one of the most abused drugs globally, but the mechanism of its addiction remains unclear. Several animal studies have shown that the gut microbiota (GM) influences addictive behaviors, but the pattern of GM changes during addiction in animals of different species remains unclear. The aim of this study was to explore the association between dynamic changes in GM and MA self-administration acquisitionamong two classical mammals, rhesus monkeys (Macaca mulatta) and rats, MA self-administration models. Male Sprague-Dawley rats and male rhesus monkeys were subjected to classical MA self-administration training, and fecal samples were collected before and after MA self-administration training, respectively. 16S rRNA sequencing was used for GM analyses. We found that GM changes were more pronounced in rats than in rhesus monkeys, as evidenced by more GM taxa producing significant differences before and after MA self-administration training in rats than in monkeys. We also found that the expression of the genus Clostridia_vadinBB60_group significantly decreased after MA self-administration training in both rats and rhesus monkeys. Lactobacillus changes were significantly negatively correlated with total MA uptake in rats (Pearson R = - 0.666, p = 0.035; Spearman R = - 0.721, p = 0.023), whereas its change was also highly negatively correlated with total MA uptake in rhesus monkeys (Pearson R = - 0.882, p = 0.118; Spearman R = - 1.000, p = 0.083), although this was not significant. These findings suggest that MA causes significant alterations in GM in both rhesus monkeys and rats and that the genus Lactobacillus might be a common therapeutic target for MA uptake prevention across the species.

Implementation of Adaptive Core Value

This research aims to describe how the Adaptive Core Value learning process can explain the Adaptive performance profile of Basic Training participants for Agenda II Civil Servant Candidates using the classical training model. The learning process for participants aims to understand the concept of Adaptiveness and actualize adaptive values in carrying out their job duties. From the concept findings and implementation, the learning process achieved will be put into practice and described; in addition to that, an instrument will be developed to see the participant's adaptive performance profile. This research used descriptive quantitative and qualitative methods with purposive random sampling involving participants from Batch I and XVII of 2022 at the Personnel and Human Resources Development Agency of Bali Province. The results of the book review found that the Adaptive Core Value concept leads to adaptive performance, which contains the dimensions: 1) handling emergency/crisis situations; 2) managing work stress; 3) solving problems creatively; 4) facing an uncertain work situation; 5) training and learning efforts; 6) interpersonal adaptation; 7) work culture adaptation; and 8) physical adaptation. The findings in this research show that all dimensions of Adaptive performance have emerged in the learning process in Adaptive training courses. In line with this, the participant's Adaptive performance profile is high. Thus, there is a match between the achievements in the learning process and the participants' adaptive performance profiles. Learning is not limited to concepts and theories but how they can actualize these Adaptive values in the form of Adaptive performance in the learning process. It is hoped that the results of this research can be a reference in learning Agenda II Core Values of ASN with AKHLAK on Adaptive values and can develop other Core Values of ASN with AKHLAK.

VR-based Competence Training at Scale: Teaching Clinical Skills in the Context of Virtual Brain Death Examination

Teaching medical practical and soft skills in clinical routines is increasingly difficult, and manikin or actor-based simulations have gained popularity in the last decades. These simulations, however, hardly scale with the demand, are commonly insufficient to train crucial clinical competencies, and cannot portray complex visual and dynamic symptomatologies as required in, for example, brain death examinations. In this paper, we explore the requirements and challenges of integrating a large-scale high-throughput VR setup into a real medical curriculum and describe our approaches and implementation. Therefore we extend and evaluate an interactive virtual reality-based simulation for training brain death diagnostics in a virtual intensive care environment, featuring a fully reactive simulated patient. To enable the required scalability we integrated the simulation into a dedicated hardware and software framework, enabling 12 simultaneous VR trainings which are controlled by a centralized server system. Using this setup we continuously collected feedback on the application's usability and realism from hundreds of students to gain first insights into the applicability of large-scale VR-based learning systems in real course designs. After integrating this feedback, we conducted a controlled curricular study in which we compared the virtual brain death simulation with the classical manikin-based training approach. Our results indicate that the immersive learning experience is perceived to be more realistic and engaging and is overall preferred by the students while also providing the same learning effect as the alternatives.

Metabolic effects of training with blood flow restriction

Relevance. Training with limited blood flow, which is performed with partial vascular occlusion using inflatable cuffs, can cause an increase in skeletal muscle mass and strength, similar to those observed with classical high-intensity training with weights, although with a noticeably lower mechanical load. Performing a load with a restriction of blood flow leads to the accumulation of metabolites, recruitment of motor neurons with a higher excitation threshold, edema of muscle cells, an increase in the concentration of reactive oxygen species, which, in turn, contributes to hypertrophy of muscle fibers, increased muscle strength, mitochondrial biogenesis, angiogenesis, increased activity of glucose transporters. Based on this, training with limited blood flow should have a positive effect on metabolic health.The aim of the study is a theoretical analysis of the metabolic effects of blood flow restriction training in healthy individuals and in clinical populations.Methods and organization of the study - this review examines the metabolic effects of TOK in both healthy and clinical populations.Conclusions. BFR can be considered as a potential method for improving muscle function and metabolism of glucose, triglycerides and fatty acids.

FMEA, monitoring, retrofit and redesign of insulated rail joints

Insulated rail joints (IRJ) are a fundamental component of railway signalling that ensure train protection using “track circuits”. A track circuit is a section of line delimited by rail joints electrically insulated, powered at one end with an electric power supply to establish a different voltage that is checked at the other end with a so-called “track circuit relay”. When the section is short-circuited by an axle the relay drops, and the section is defined as “occupied”. IRJ are subjected to continuous and dangerous failures as they represent a singularity in the rail continuity. Continuous battering produces the classical train impact noise while damage in the rail accumulates until the IRJ fails with extremely unpleasant consequences. The paper discusses the main failure modes of an IRJ starting from a historical perspective, showing how an appropriate failure analysis would have prevented a catastrophic accident that recently occurred in Italy. Several countermeasures are shown and a completely new IRJ potentially completely solving the issue is introduced.

Prediction of monkeypox infection from clinical symptoms with adaptive artificial bee colony-based artificial neural network

In 2022, the World Health Organization declared an outbreak of monkeypox, a viral zoonotic disease. With time, the number of infections with this disease began to increase in most countries. A human can contract monkeypox by direct contact with an infected human, or even by contact with animals. In this paper, a diagnostic model for early detection of monkeypox infection based on artificial intelligence methods is proposed. The proposed method is based on training the artificial neural network (ANN) with the adaptive artificial bee colony algorithm for the classification problem. In the study, the ABC algorithm was preferred instead of classical training algorithms for ANN because of its effectiveness in numerical optimization problem solutions. The ABC algorithm consists of food and limit parameters and three procedures: employed, onlooker and scout bee. In the algorithm standard, artificial onlooker bees are produced as much as the number of artificially employed bees and an equal number of limit values are assigned for all food sources. In the advanced adaptive design, different numbers of artificial onlooker bees are used in each cycle, and the limit numbers are updated. For effective exploitation, onlooker bees tend toward more successful solutions than the average fitness value of the solutions, and limit numbers are updated according to the fitness values of the solutions for efficient exploration. The performance of the proposed method was investigated on CEC 2019 test suites as examples of numerical optimization problems. Then, the system was trained and tested on a dataset representing the clinical symptoms of monkeypox infection. The dataset consists of 240 suspected cases, 120 of which are infected and 120 typical cases. The proposed model's results were compared with those of ten other machine learning models trained on the same dataset. The deep learning model achieved the best result with an accuracy of 75%. It was followed by the random forest model with an accuracy of 71.1%, while the proposed model came third with an accuracy of 71%.

Defect Detection of Cell Phone Screen Using a Faster Regional Convolutional Neural Network with Multi-head Attention Mechanism

Since the glass outer screen of a cell phone is the main sensory part of the human eye when using a cell phone, the advantages and disadvantages of the cell phone screen directly affect people's sense of use. Therefore, the defect detection requirements for cell phone screens are high and need to meet the needs of high-volume factory inspection. Most of the traditional defect detection methods use visual methods, the detection results are overly dependent on the subjectivity and experience of workers, the efficiency of this method is low, and the accuracy is poor. Currently, machine learning-based detection methods are applied in numerous industries. In this paper, a faster Regional Convolutional Neural Network (R-CNN) with multi-head attention mechanism for defect detection of cell phone screen is proposed. To enhance the network's capability in extracting feature information, a four-head attention mechanism is added to the last convolutional layer of the ResNet50 network. An improved Region of Interest (ROI) Align is proposed to replace the original ROI Pooling to reduce the localization error of cell phone screen defects. Replace the original Rectified Linear Unit (ReLU) activation function with the Copy Exponential Linear Unit (CELU) activation function to expedite the convergence capability of the network. Finally, by comparing with other classical model training, the evaluation results indicate that the proposed method achieved an average accuracy of 95.71%, which is a 5.34% improvement compared to the original faster R-CNN network.

Distributed entropy-regularized multi-agent reinforcement learning with policy consensus

Sample efficiency is a limiting factor for existing distributed multi-agent reinforcement learning (MARL) algorithms over networked multi-agent systems. In this paper, the sample efficiency problem is tackled by formally incorporating the entropy regularization into the distributed MARL algorithm design. Firstly, a new entropy-regularized MARL problem is formulated under the model of networked multi-agent Markov decision processes with observation-based policies and homogeneous agents, where the policy parameter sharing among the agents provably preserves the optimality. Secondly, an on-policy distributed actor–critic algorithm is proposed, where each agent shares its parameters of both the critic and actor for consensus update. Then, the convergence analysis of the proposed algorithm is provided based on the stochastic approximation theory under the assumption of linear function approximation of the critic. Furthermore, a practical off-policy version of the proposed algorithm is developed which possesses scalability, data efficiency and learning stability. Finally, the proposed distributed algorithm is compared against the solid baselines including two classic centralized training algorithms in the multi-agent particle environment, whose learning performance is empirically demonstrated through extensive simulation experiments.