Under the background of strategic emerging professional construction,the teaching reform of the course Sensor Principles and Detection Technology in the Robotics Engineering major is of great significance.This paper ex...Under the background of strategic emerging professional construction,the teaching reform of the course Sensor Principles and Detection Technology in the Robotics Engineering major is of great significance.This paper explores the current challenges faced in the teaching of this course,such as the rapid development of sensor technology,the increasing demand for practical skills in the roboticsfield,and the need to integrate interdisciplinary knowledge.To address these challenges,several reform measures are proposed.Firstly,the curriculum content is updated to include the latest developments in sensor technology and its applications in robotics.This involves integrating advanced topics such as intelligent sensors,MEMS technology,and wireless sensor networks into the syllabus.Secondly,the teaching methods are diversified.Interactive teaching methods,such as project-based learning and case-based learning,are adopted to enhance students practical abilities and problem-solving skills.In addition,the use of virtual simulation software and laboratory experiments is emphasized to provide students with hands-on experience in sensor design,testing,and integration.Thirdly,the assessment system is reformed to focus more on students comprehensive abilities.In addition to traditional exams,project reports,laboratory performance,and team collaboration are also included in the assessment criteria.The implementation of these reform measures has achieved positive results.Students interest in the course has increased,their understanding of sensor principles and detection technology has been deepened,and their practical skills and innovative thinking have been significantly improved.This reform provides valuable references for the teaching of similar courses in other engineering majors and contributes to the cultivation of high-quality talents in thefield of robotics engineering to meet the needs of strategic emerging industries.展开更多
In the current era of rapid technological development,robot technology has become a key forcedriving the transformation of various industries.Applied undergraduate education shoulders the important responsibilityof cu...In the current era of rapid technological development,robot technology has become a key forcedriving the transformation of various industries.Applied undergraduate education shoulders the important responsibilityof cultivating high-quality robot professionals who can meet the demands of industries.This studyfocuses on constructing a robot course teaching model based on the STEAM concept in application-orientedundergraduate colleges and conducts an in-depth exploration of its application effect,with the expectation ofproviding useful references for the teaching reform of related courses.Firstly,the article deeply analyzes theconnotation of the STEAM concept,namely interdisciplinary integration(science,technology,engineering,art,mathematics),practical innovation and the cultivation of comprehensive quality,emphasizing its fit andimportance for the teaching of robot courses.Robotics technology itself is a field that integrates multiple disciplines,covering knowledge from various aspects such as mechanical design,electronic engineering,computerprogramming,and sensor technology.At the same time,it also requires students to possess innovative thinkingand practical abilities to solve complex problems in practical applications,which is highly consistent with theeducational goals advocated by the STEAM concept.It has laid a solid theoretical foundation for the constructionof the teaching mode.Secondly,the specific construction process of the teaching mode was elaborated indetail.In terms of the curriculum system,it breaks through the limitations of the traditional single-disciplinecurriculum setting and builds a modular curriculum system covering multi-disciplinary knowledge,such as settingup basic modules(mathematical modeling,physics and mechanics,etc.),professional core modules(robotprogramming,mechanical structure design,etc.),and extension modules(artificial intelligence and robots,robot art design,etc.).Enable students to systematically learn knowledge related to robots and achieve theorganic integration and expansion of knowledge.In terms of teaching methods,a variety of methods such asproject-based teaching,problem-oriented teaching and group cooperative learning are comprehensively applied.Take project-based teaching as an example.Teachers assign robot project tasks with practical applicationbackgrounds,such as designing a robot for campus environment monitoring.Students need to independentlyconsult materials,form teams,and collaborate in division of labor.They participate throughout the entireprocess from the overall conception of the robot,mechanical structure construction,circuit design,programwriting to the final debugging and operation.This not only exercises students’practical hands-on ability,butalso cultivates their comprehensive ability to solve complex problems and teamwork spirit.In the teaching evaluationlink,a diversified evaluation system has been constructed.It has changed the previous evaluation methodmainly based on examination scores,combined process evaluation with outcome evaluation,and focused onconducting comprehensive evaluations of students’performance during the learning process,teamwork ability,innovation ability,and the quality of project completion,etc.By observing students’performance during theproject implementation process,mutual evaluation among group members,project achievement presentationand defense,etc.,comprehensively and objectively evaluate students’learning outcomes,promptly identifytheir strengths and weaknesses,and provide a basis for subsequent teaching improvement.Finally,throughthe teaching practice application carried out in the robotics major of an application-oriented undergraduatecollege,the constructed teaching mode was verified.The practical results show that students’enthusiasm andinitiative in the robot course learning have significantly increased,and their classroom participation has greatlyimproved.They have shifted from passively accepting knowledge in the past to actively exploring and practicing.In terms of the achievements of course projects,the robot works designed by students are more innovativeand practical,and can well meet the actual application needs.For example,some intelligent logistics robots designedby students have conducted preliminary application tests in the campus express delivery sorting scenarioand achieved good results.Meanwhile,the number of awards won by students in various robot competitionshas also significantly increased,which fully demonstrates the improvement of students’comprehensive qualityand practical innovation ability,and strongly proves the effectiveness and feasibility of the robot course teachingmodel based on the STEAM concept.In conclusion,the application-oriented undergraduate robot courseteaching model based on the STEAM concept constructed in this study has effectively promoted students’interdisciplinary knowledge integration,practical innovation and comprehensive quality improvement by optimizingthe curriculum system,innovating teaching methods and improving the teaching evaluation system.Itprovides new ideas and methods for the cultivation of robot professionals in application-oriented undergraduatecolleges and universities.It has significant theoretical and practical significance,and can provide reference and inspiration for the teaching reform of other related professional courses,promoting applied undergraduateeducation to better adapt to social development and industrial demands.展开更多
The Outcome-Based Education(OBE)philosophy has been widely recognized as an effective ap-proach to enhance the quality of engineering education.This paper focuses on the construction of a talent cultivation system for...The Outcome-Based Education(OBE)philosophy has been widely recognized as an effective ap-proach to enhance the quality of engineering education.This paper focuses on the construction of a talent cultivation system for the Robotics Engineering major based on the OBE concept.The OBE approach em-phasizes the importance of clearly defining expected learning outcomes and aligning the entire educational process to achieve these outcomes.In the context of Robotics Engineering,this means identifying the specific skills,knowledge,and competencies that graduates should possess to meet the demands of the industry and society.Firstly,the paper analyzes the current challenges and needs in thefield of Robotics Engineering.It identifies the key areas where graduates need to excel,such as programming,mechanical design,electronics,and artificial intelligence.Based on this analysis,the expected learning outcomes for the Robotics Engineering major are defined.These outcomes cover not only technical skills but also soft skills such as problem-solving,teamwork,and innovation.The paper then discusses the curriculum design that is aligned with these learning outcomes.The curriculum includes a combination of theoretical courses and practical training.The theoretical courses provide students with a solid foundation in the principles of robotics,while the practical training allows students to apply their knowledge to real-world problems.The curriculum also incorporates project-based learning,where students work on projects that simulate real-world scenarios.This approach helps students to develop their problem-solving skills and gain hands-on experience.In addition to the curriculum design,the paper also addresses the assessment methods.The assessment is outcome-based,meaning that it focuses on evaluating whether students have achieved the expected learning outcomes.This includes both formative and summative assessments,such as quizzes,exams,project evaluations,andfinal capstone projects.The as-sessment results are used to provide feedback to students and to continuously improve the teaching process.Finally,the paper presents the implementation of the talent cultivation system and evaluates its effectiveness.The results show that the OBE-based system has significantly improved the quality of education in the Robotics Engineering major.Graduates from this system are better prepared to meet the challenges of the industry and have a higher employability rate.In conclusion,the construction of a talent cultivation system for Robotics Engineering based on the OBE philosophy is an effective way to improve the quality of engineering educa-tion.By clearly defining learning outcomes and aligning the curriculum and assessment methods to achieve these outcomes,this system helps to produce well-rounded graduates who are ready to contribute to thefield.展开更多
Traditional deep learning-based shadow removal methods often alter the pixels in non-shadow areas and fail to produce shadow removal results with natural boundary transitions.To solve this problem,we propose a novel m...Traditional deep learning-based shadow removal methods often alter the pixels in non-shadow areas and fail to produce shadow removal results with natural boundary transitions.To solve this problem,we propose a novel multi-stage shadow removal framework based on Generative Adversarial Networks(GAN).Firstly,the multi-task-driven generator respectively generates the corresponding shadow mask and shadow mask for the input image through the shadow detection subnet and the mask generation subnet.Secondly,guided by the shadow mask and shadow mask,the full shadow module and the partial shadow module are respectively de-signed,and different types of shadows in the image are removed in stages.The sequential data from multiple sensors is used as the input of the time series generative adversarial network to generate sequence data with temporal dynamic characteristics;the data synthesized by the time series generative adversarial network is used to replace the noise input data in the gradient-penalized WGAN generative adversarial network,and the discriminator combines graph convolutional networks,long short-term memory networks and attention mech-anisms to more effectively explore the spatio-temporal correlations of multi-source heterogeneous sensing data and enhance the discriminative ability for sequential data.Then,a new combined loss function was constructed based on the least squares loss to achieve better results.Compared with the latest deep learning shadow removal methods,on the selected dataset,the balance error rate of the proposed method decreases by approximately 4.39%,the structural similarity increased by approximately 0.44%,and the pixel root mean square error de-creases by approximately 13.32%.The experimental results show that the shadow removal results obtained by this method have smoother boundary transitions.展开更多
Under the backdrop of the comprehensive advancement of new engineering education,the robotics engineering discipline,which integrates multiple fields such as mechanics,electronics,control,computer science,and artifici...Under the backdrop of the comprehensive advancement of new engineering education,the robotics engineering discipline,which integrates multiple fields such as mechanics,electronics,control,computer science,and artificial intelligence,urgently needs to break through traditional disciplinary barriers to meet the pressing demand for interdisciplinary talents in the era of intelligent manufacturing and artificial intelligence.This paper systematically reviews the core issues existing in the current course construction of the robotics engineering major,such as a monotonous course structure,weak practical components,and insufficient interdisciplinary integration.It proposes a course system reconstruction path based on the concept of solid foundation,strong cross-disciplinary integration,emphasis on practice,and promotion of innovation.The research emphasizes that a flexible course structure of platform+module+direction should be established around industrial demands and technological frontiers,strengthening the mathematical and physical foundation and core professional courses,and expanding courses in cutting-edge areas such as artificial intelligence,big data,and cloud-edge collaboration.It also advocates for the reform of practical teaching models through project-driven,competition-education integration,and school-enterprise collaboration.Meanwhile,ideological and political education,innovation and entrepreneurship education,and engineering ethics should be organically integrated into the entire course process to form a trinity talent cultivation pattern of value shaping,knowledge imparting,and ability development.By constructing a multi-level,three-dimensional,and open course system for the robotics engineering major,it can provide strong support for the cultivation of high-level applied and innovative talents under the new engineering background,and is of great significance for promoting the connotative development of higher education and the high-quality development of industries.展开更多
Artificial intelligence(AI)is profoundly reshaping the practical logic and theoretical boundaries of bel canto education.This study takes“technical path-aesthetic challenge-paradigm reconstruction”as the analytical ...Artificial intelligence(AI)is profoundly reshaping the practical logic and theoretical boundaries of bel canto education.This study takes“technical path-aesthetic challenge-paradigm reconstruction”as the analytical framework to systematically explore the innovative mechanism and practical path of AI-empowered vocal education.At the technical level,an acoustic parameter modeling system is constructed based on deep learning.Through MEL spectrum analysis,forresonance tracking and breath dynamics modeling,the traditional singing assessment that relies on auditory experience is transformed into quantifiable and interpretable acoustic indicators(such as harmonic energy ratio,vowel resonance stability,etc.),achieving precise diagnosis and real-time intervention of vocal defects.Experimental data shows that this system reduces the learners’pitch error rate by 42%and increases their breath control efficiency by 35%.In the dimension of personalized training,an adaptive teaching engine based on reinforcement learning was developed to dynamically generate a three-dimensional matching scheme of“voice features-track difficulty-training intensity”,significantly reducing the time for beginners to master core vocal techniques(by an average of 28%).However,the conflict between AI quantitative indicators and the traditional aesthetic standards of vocal music has become prominent:the algorithm’s preference for standardized formuster distribution may suppress the singer’s unique timbre personality and artistic expression tension.To this end,a“dual-track evaluation model”is proposed:taking acoustic parameters as the technical benchmark and emotional expression and artistic appeal as the aesthetic benchmark,and achieving a dynamic balance between the two through expert annotation and group consensus algorithms.The research further reconstructs the teaching paradigm,advocating that AI tools be positioned as“aesthetic collaborators”rather than substitutes,and builds a new type of teacher-student relationship of“human-machine co-teaching-co-evaluation-co-creation”.This research provides a solution that is both technically feasible and aesthetically reasonable for the digital transformation of bel canto education,revealing the underlying logic of the deep integration of art and technology in the AI era.展开更多
Hierarchical clustering is a widely used technique for data grouping in various fields,renowned for its ability to create a dendrogram that provides insights into data structure.However,traditional hierarchical method...Hierarchical clustering is a widely used technique for data grouping in various fields,renowned for its ability to create a dendrogram that provides insights into data structure.However,traditional hierarchical methods often suffer from challenges such as computational inefficiency,sensitivity to noise,and difficulty in determining the optimal number of clusters.This paper introduces an improved hierarchical clustering algorithm that incorporates advanced distance metrics,the use of cluster representatives,and a robust agglomeration strategy designed to address these challenges.The proposed algorithm incorporates a mechanism to dynamically select the distance metric based on the underlying data characteristics.Unlike traditional methods that rely solely on a fixed distance metric,our approach determines the most suitable metric for the given dataset.For instance,a method like Cosine similarity could be employed for high-dimensional data or text data where orientation matters,while Euclidean distance can be effective for low-dimensional,continuous data.Our proposed method is evaluated through comprehensive experiments on synthetic and real-world datasets,showcasing significant enhancements in clustering performance,adaptability to noise,and computational efficiency.The findings indicate that the proposed algorithm outperforms traditional hierarchical methods,demonstrating its potential for broader applications across various domains.展开更多
Although Chinese-English machine translation is a resource-rich language pair,the problem of data sparsity still exists.For example,for the translation of some specific domains or low-frequency words,the amount of par...Although Chinese-English machine translation is a resource-rich language pair,the problem of data sparsity still exists.For example,for the translation of some specific domains or low-frequency words,the amount of parallel corpus is limited,which makes it difficult to improve the translation quality of the model in these scenarios.Neural machine translation models usually need a large amount of alignment data to train,otherwise they are prone to over-fitting.Neural machine translation models are slow to train and decode,especially when dealing with long sentences or complex structures,which limits their efficiency in real-time application scenarios.Chinese-English machine translation can help people overcome the language barrier and promote the communication and cooperation between people with different language backgrounds,which is of great significance for international business,academic exchanges and cultural exchanges.This paper proposes a Chinese-English neural machine translation model based on generative adversarial network.This model applies the generative adversarial network to neural machine translation,and further optimizes the adversarial learning based neural machine translation model by improving the monotone decoding sequence from left to right or from right to left in the original machine translation model.At the same time,unlike previous generative adversarial networks,neural machine translation models are actually a sequence of discrete symbols that map source language sentences to target language sentences,both in discontinuous spaces.In this case,the generative adversarial network fails to transmit the gradient properly,causing the generator to lose its update direction.By introducing the strategy gradient algorithm in reinforcement learning,the generator optimization problem in adversarial learning is solved,and the translation performance of the model is improved.Finally,experiments on public data sets show that the proposed model can effectively improve translation quality compared with other advanced models.展开更多
In the new era,vocational college has been an important component of the education system in China,responsible for the mission of cultivating high-quality technical talents.The construction of one-stop student communi...In the new era,vocational college has been an important component of the education system in China,responsible for the mission of cultivating high-quality technical talents.The construction of one-stop student communities has become an important way to promote the modernization of vocational education governance.With the deepening of education reform,vocational colleges are facing new challenges in the con-struction of student communities.The contradiction between the expansion of the scale of vocational college students and the diversification of demand is prominent,and the traditional management model has shortcom-ings such as scattered resources and insufficient collaboration.How to explore a one-stop student community construction path that is in line with the characteristics of vocational colleges has become an urgent issue to be solved.Nowadays,the development of artificial intelligence(AI)technology is flourishing,and deeply changing the face of higher education.The rapid development of AI technology provides strong support for the construction of student communities in vocational colleges.AI intelligent tools,big data information col-lection and other technologies,can effectively help colleges continuously optimize their management models,greatly improve service efficiency and quality,and enhance student satisfaction.Hence,based on the AI tech-nology,building an equal,inclusive,and interactive student community network,responding to students’needs for digital learning resources,platform tools,service support,and improving the accuracy and satisfaction of education management,are of great significance for the vocational education reform in China.This article overviews pervious achievements of AI-enabled student community construction in colleges,to provide ac-tionable solutions for the paradigm shift of the student service system in vocational colleges at the practical level.展开更多
In the pedestrian detection scenario,the detection algorithm usually misses obscured and distant fuzzy pedestrians,and at the same time cannot take into account the detection accuracy and speed.In this paper,we propos...In the pedestrian detection scenario,the detection algorithm usually misses obscured and distant fuzzy pedestrians,and at the same time cannot take into account the detection accuracy and speed.In this paper,we propose a modified YOLOv5 model for pedestrian detection.Firstly,the backbone network uses the SPD-GCONV module constructed by the combination of SPD(Space-to-Depth)module and Ghost convolution for down-sampling to reduce the loss of fine-grained feature information.Secondly,the multi-scale detection ability of the model is enhanced by adding a small size detection layer.Then,the original CIoU loss function is replaced by α-EloU loss function to improve the accuracy of pedestrian target location.According to the experiments on WiderPerson data set,the average detection accuracy is improved by 2%compared with other pedestrian detection algorithms on the premise of ensuring the detection speed.Experimental results show that the improved algorithm can significantly improve the detection performance.展开更多
Fourth party logistics(4PL)supply chain network design problem considering the behavior of customer service satisfaction becomes more important in supply chain network design problem.It is NP-hard problem.Improved dif...Fourth party logistics(4PL)supply chain network design problem considering the behavior of customer service satisfaction becomes more important in supply chain network design problem.It is NP-hard problem.Improved differential evolution(IDE)algorithm is applied to solve this problem in this paper.In order to handle the infeasible solution efficiently,a two-stage algorithm,the minimum cost flow incorporated with macro-micro adjustment embedded improved differential evolution(MMAMFC-IDE),is designed.The results of numerical experiments show that MMAMFC-IDE can get relatively better results compared with the minimum cost flow embedded improved differential evolution(MFC-IDE).展开更多
Industry-education integration communities are redefining how advanced robotics courses are designed,delivered and validated.This paper reports a two-year design-based study that reconceptualised the undergraduate mod...Industry-education integration communities are redefining how advanced robotics courses are designed,delivered and validated.This paper reports a two-year design-based study that reconceptualised the undergraduate module Robot Vision Perception and Detection from a supply-chain-oriented alliance of three universities,two robotics manufacturers and one logistics giant.Adopting a community-of-practice lens,we replaced the conventional lectureClab sequence with a challenge-driven co-creation loop in which corporate engineers release real-time production-line vision defects as curricular tasks,faculty scaffold theoretical principles,and students iterate solutions on the factory floor using identical hardware and data streams.Mixedmethods evaluation with 142 students and 18 industry mentors shows significant gains:(1)learning performance increased by 0.82 standard deviations;(2)student creative self-efficacy and systems-thinking improved 34%and 29%respectively;(3)average defect detection recall of student models rose from 72%to 93%,with 8 prototypes transferred to the partner lines;(4)facultyCindustry co-publications and patent disclosures tripled.Qualitative trace data reveal that boundary objects(annotated datasets,Dockerized algorithms and shared Kanban boards)legitimately brokered epistemic differences between academia and industry.The study contributes an empirically grounded frameworkłCIE-CDL(Community-Integrated Education via Challenge-Driven Loops)łthat embeds authentic socio-technical complexity into robotics curricula while simultaneously generating measurable value for industrial partners.Implications for scalable,sustainable industryCeducation symbiosis are discussed.展开更多
The rapid development of artificial intelligence technology has provided an opportunity to reshape the teaching ecosystem in open education.This article focuses on the concept of“adaptive learning”,in the context of ...The rapid development of artificial intelligence technology has provided an opportunity to reshape the teaching ecosystem in open education.This article focuses on the concept of“adaptive learning”,in the context of the artificial intelligence era,and explores the systematic reform of open education teaching models.The researchfirst constructed an integrated learning framework that combines cognitive diagnosis,dynamic paths,resource push,immediate feedback,and emotional support.Through data-driven and teacher experience collaboration,it realizes large-scale personalized teaching.Secondly,based on the teaching practice of public courses in multiple universities,the article collected and analyzed the entire process behavior data of learners,used deep models to dynamically optimize teaching strategies,and established an interpretable and iterative teaching loop.On this basis,the research focuses on educational equity and the mechanism of human-computer collaboration,ensuring that while technology is empowered,the dominant position of teachers and the warmth of the learning community are maintained.Through qualitative interviews and teaching observations,the article found that adaptive learning significantly enhanced the initiative,satisfaction,and knowledge transfer ability of learners,forming a new classroom culture that integrates online and offline elements and reshapes the roles of teachers and students.The research conclusion states that in the open education teaching reform of the artificial intelligence era,it should be driven by data intelligence,centered on learners,and based on educational equity,promoting the transformation from“standardized supply”to“precise services”,providing replicable models and sustainable paths for building a lifelong learning society.展开更多
In recent years,neural machine translation has rapidly developed and replaced traditional machine translation,becoming the mainstream paradigm in the field of machine translation.Machine translation can reduce transla...In recent years,neural machine translation has rapidly developed and replaced traditional machine translation,becoming the mainstream paradigm in the field of machine translation.Machine translation can reduce translation costs and improve translation efficiency,bring good news to cultural exchanges and international cooperation,and help national development.However,neural machine translation is highly dependent on large-scale high-quality parallel corpus,and there are problems such as uneven quality and sparse data,so it is imperative to study and explore neural machine translation.The purpose of this paper is to construct pseudo-parallel corpus using data enhancement technology,improve the diversity of Chinese and English materials,and then optimize the translation model to improve the translation effect of the model.Based on BERT pre-training technology,this paper first analyzes the limitations of the traditional Transformer model,and then puts forward two directions for model optimization.On the one hand,in the data preprocessing stage,multi-granularity word segmentation technology is used for word segmentation to help Chinese-English neural machine translation model better understand the text.On the other hand,in the pre-training stage,this paper adopts the strategy of deep integration of BERT dynamic word embedding and original word embedding.On the basis of the original Transformer,a fusion module is added,through which the original word embeddings and BERT dynamic word embeddings are simple linear splicing,and then fed into the encoder.The attention mechanism is used for deep integration and better word vector representation,enabling the Transformer model to take full advantage of the external semantic information introduced by BERT.Finally,the feasibility and effectiveness of the Transformer architecture adopted in this paper are verified by the comparison experiment between RNN and Transformer model.Through the ablation experiments of different word vector representation and different stages using BERT pre-training technology,the effectiveness of BERT dynamic word embedding and deep fusion of word embedding and the rationality of using pre-training technology only in the encoder stage are verified.展开更多
This paper proposes a novel image enhancement based on guidedfiltering and modified Retinex method to address the problems of poor visual effect,unclear images and incomplete display of details in low-light images.Low-i...This paper proposes a novel image enhancement based on guidedfiltering and modified Retinex method to address the problems of poor visual effect,unclear images and incomplete display of details in low-light images.Low-illumination images are processed using guidedfiltering to adaptively enhance the luminance component.This method consists of two parts:the image decomposition network and the image en-hancement network.The image decomposition network is responsible for decomposing the original image into illumination components and reflection components.The image enhancement network optimizes parameters and performsγcorrection through the natural image quality evaluator(NIQE),adjusting the brightness and contrast of the illumination component,and then re-fuses it with the reflection component to improve the over-all image quality.Finally,image quality enhancement experiments are conducted using the images in the SOTS standard test set and six actual scene images.They are compared and analyzed with other advanced methods.The experimental results show that the proposed method is superior to other methods in both subjective visual effects and objective evaluation indicators,thus fully verifying the effectiveness and feasibility of the proposed method.展开更多
Due to the large scale and high dimension of teaching data,the using of traditional clustering algorithms has problems such as high computational complexity and low accuracy.Therefore,this paper proposes a weighted bl...Due to the large scale and high dimension of teaching data,the using of traditional clustering algorithms has problems such as high computational complexity and low accuracy.Therefore,this paper proposes a weighted block sparse subspace clustering algorithm based on information entropy.The introduction of information entropy weight and block diagonal constraints can obtain the prior probability that two pixels belong to the same category before the simulation experiment,thereby positively intervening that the solutions solved by the model tend to be the optimal approximate solutions of the block diagonal structure.It can enable the model to obtain the performance against noise and outliers,and thereby improving the discriminative ability of the model classification.The experimental results show that the average probability Rand index of the proposed method is 0.86,which is higher than that of other algorithms.The average information change index of the proposed method is 1.55,which is lower than that of other algorithms,proving its strong robustness.On different datasets,the misclassification rates of the design method are 1.2%and 0.9%respectively,which proves that its classification accuracy is relatively high.The proposed method has high reliability in processing highdimensional teaching data.It can play an important role in the field of educational data analysis and provide strong support for intelligent teaching.展开更多
Deep learning is more and more widely used in natural language processing.Compared with the traditional n-gram statistical language model,Recurrent neural network(RNN)modeling technology has shown great advantages in ...Deep learning is more and more widely used in natural language processing.Compared with the traditional n-gram statistical language model,Recurrent neural network(RNN)modeling technology has shown great advantages in language modeling,and has been gradually applied in speech recognition,machine translation and other fields.However,at present,the training of RNN language models is mostly offline.For different speech recognition tasks,there are language differences between training corpus and recognition tasks,which affects the recognition rate of speech recognition systems.While using RNN modeling technology to train the Chinese language model,an online RNN model self-adaption algorithm is proposed,which takes the preliminary recognition results of speech signals as corpus to continue training the model,so that the adaptive RNN model can get the maximum match with the recognition task.The experimental results show that the adaptive model effectively reduces the language difference between the language model and the recognition task,and the recognition rate of the system is further improved after the Chinese word confusion network is re-scored,which has been verified in the actual Chinese speech recognition system.展开更多
This study designs a curriculum system that deeply integrates industry and education for robotics engineering majors in applied universities.Guided by the CDIO philosophy and the concept of industryCeducation collabor...This study designs a curriculum system that deeply integrates industry and education for robotics engineering majors in applied universities.Guided by the CDIO philosophy and the concept of industryCeducation collaboration,a three-level curriculum framework consisting of foundational courses,professional cores,and integrated practical modules is constructed.Industry technologies and standards are introduced into course content,teaching methods,and evaluation systems through co-developed courses,shared laboratories,and joint faculty teams.The system emphasizes continuous engagement with real-world engineering projects,fostering students’competencies in design,programming,integration,and innovation.The proposed model provides a flexible and sustainable approach for cultivating application-oriented robotics talents that meet the evolving needs of intelligent manufacturing.展开更多
This paper proposes a novel arrhythmia classification method combining Convolutional Neural Networks(CNN),Attention mechanisms,and Bidirectional Transformers(BiTransformer).The method aims to improve the accuracy and ...This paper proposes a novel arrhythmia classification method combining Convolutional Neural Networks(CNN),Attention mechanisms,and Bidirectional Transformers(BiTransformer).The method aims to improve the accuracy and robustness of arrhythmia detection in electrocardiogram(ECG)signals.Initially,the CNN module extracts local spatial features from raw ECG data,effectively capturing the morphological characteristics of different arrhythmia types.Subsequently,the Attention mechanism is applied to weigh the importance of different segments in the ECG signal,allowing the model to focus on critical features that are most indicative of arrhythmia.Finally,the BiTransformer module processes the feature sequences bidirectionally,capturing both forward and backward dependencies in the ECG signal.This comprehensive approach enables the model to integrate local and global information,enhancing its ability to classify various arrhythmias accurately.Experiments conducted on the MIT-BIH Arrhythmia Database demonstrate that the proposed method achieves state-of-the-art performance,with a significant improvement in classification accuracy compared to traditional methods.The results highlight the effectiveness of combining CNN,Attention,and BiTransformer for arrhythmia classification,offering a promising direction for automated ECG analysis and clinical applications.展开更多
This study is based on the theory of human engineering psychology and explores the mechanism by which cool and warm color posters affect the emotions of college students.Through the combination of exper-imental method...This study is based on the theory of human engineering psychology and explores the mechanism by which cool and warm color posters affect the emotions of college students.Through the combination of exper-imental methods and scale measurement,60 college students were selected as participants.Cool color(blue,green)and warm color(red,yellow)posters were randomly presented to them,and their emotional responses were collected.The SAM scale(Self-Assessment Manikin)was used to assess the degree of pleasure,arousal,and dominance,and skin conductance response(GSR)was used to obtain physiological indicators to enhance the objectivity of the study.The results showed that warm color posters significantly enhanced the arousal and pleasure of college students,especially under the red stimulus,with the highest level of emotional activation.While cool color posters significantly reduced arousal and brought a more calm and relaxed emotional expe-rience.The blue tone was the most effective in alleviating anxiety.Gender and professional background have a moderating effect on color perception.Women are more sensitive to the emotional responses to cool and warm colors,and students majoring in art design have a higher recognition and emotional response intensity of color differences than non-art majors.This study verified the applicability of color emotional effects in the col-lege student population and expanded the application boundaries of human engineering psychology in visual communication design.The research suggests that in college mental health publicity and campus environment design,scientific application of color psychology principles should be adopted,and appropriate colors should be selected according to different emotional regulation goals to improve the accuracy and effectiveness of emo-tional intervention.In the future,neuroscientific methods such as electroencephalogram(EEG)can be further introduced to deeply reveal the neural basis of the color emotional mechanism.展开更多
文摘Under the background of strategic emerging professional construction,the teaching reform of the course Sensor Principles and Detection Technology in the Robotics Engineering major is of great significance.This paper explores the current challenges faced in the teaching of this course,such as the rapid development of sensor technology,the increasing demand for practical skills in the roboticsfield,and the need to integrate interdisciplinary knowledge.To address these challenges,several reform measures are proposed.Firstly,the curriculum content is updated to include the latest developments in sensor technology and its applications in robotics.This involves integrating advanced topics such as intelligent sensors,MEMS technology,and wireless sensor networks into the syllabus.Secondly,the teaching methods are diversified.Interactive teaching methods,such as project-based learning and case-based learning,are adopted to enhance students practical abilities and problem-solving skills.In addition,the use of virtual simulation software and laboratory experiments is emphasized to provide students with hands-on experience in sensor design,testing,and integration.Thirdly,the assessment system is reformed to focus more on students comprehensive abilities.In addition to traditional exams,project reports,laboratory performance,and team collaboration are also included in the assessment criteria.The implementation of these reform measures has achieved positive results.Students interest in the course has increased,their understanding of sensor principles and detection technology has been deepened,and their practical skills and innovative thinking have been significantly improved.This reform provides valuable references for the teaching of similar courses in other engineering majors and contributes to the cultivation of high-quality talents in thefield of robotics engineering to meet the needs of strategic emerging industries.
基金supported by the Project Name:Construction and Application of Teaching Mode for Applied Undergraduate Robotics Courses under the STEAM Concept Project Number:2024JGYB15.
文摘In the current era of rapid technological development,robot technology has become a key forcedriving the transformation of various industries.Applied undergraduate education shoulders the important responsibilityof cultivating high-quality robot professionals who can meet the demands of industries.This studyfocuses on constructing a robot course teaching model based on the STEAM concept in application-orientedundergraduate colleges and conducts an in-depth exploration of its application effect,with the expectation ofproviding useful references for the teaching reform of related courses.Firstly,the article deeply analyzes theconnotation of the STEAM concept,namely interdisciplinary integration(science,technology,engineering,art,mathematics),practical innovation and the cultivation of comprehensive quality,emphasizing its fit andimportance for the teaching of robot courses.Robotics technology itself is a field that integrates multiple disciplines,covering knowledge from various aspects such as mechanical design,electronic engineering,computerprogramming,and sensor technology.At the same time,it also requires students to possess innovative thinkingand practical abilities to solve complex problems in practical applications,which is highly consistent with theeducational goals advocated by the STEAM concept.It has laid a solid theoretical foundation for the constructionof the teaching mode.Secondly,the specific construction process of the teaching mode was elaborated indetail.In terms of the curriculum system,it breaks through the limitations of the traditional single-disciplinecurriculum setting and builds a modular curriculum system covering multi-disciplinary knowledge,such as settingup basic modules(mathematical modeling,physics and mechanics,etc.),professional core modules(robotprogramming,mechanical structure design,etc.),and extension modules(artificial intelligence and robots,robot art design,etc.).Enable students to systematically learn knowledge related to robots and achieve theorganic integration and expansion of knowledge.In terms of teaching methods,a variety of methods such asproject-based teaching,problem-oriented teaching and group cooperative learning are comprehensively applied.Take project-based teaching as an example.Teachers assign robot project tasks with practical applicationbackgrounds,such as designing a robot for campus environment monitoring.Students need to independentlyconsult materials,form teams,and collaborate in division of labor.They participate throughout the entireprocess from the overall conception of the robot,mechanical structure construction,circuit design,programwriting to the final debugging and operation.This not only exercises students’practical hands-on ability,butalso cultivates their comprehensive ability to solve complex problems and teamwork spirit.In the teaching evaluationlink,a diversified evaluation system has been constructed.It has changed the previous evaluation methodmainly based on examination scores,combined process evaluation with outcome evaluation,and focused onconducting comprehensive evaluations of students’performance during the learning process,teamwork ability,innovation ability,and the quality of project completion,etc.By observing students’performance during theproject implementation process,mutual evaluation among group members,project achievement presentationand defense,etc.,comprehensively and objectively evaluate students’learning outcomes,promptly identifytheir strengths and weaknesses,and provide a basis for subsequent teaching improvement.Finally,throughthe teaching practice application carried out in the robotics major of an application-oriented undergraduatecollege,the constructed teaching mode was verified.The practical results show that students’enthusiasm andinitiative in the robot course learning have significantly increased,and their classroom participation has greatlyimproved.They have shifted from passively accepting knowledge in the past to actively exploring and practicing.In terms of the achievements of course projects,the robot works designed by students are more innovativeand practical,and can well meet the actual application needs.For example,some intelligent logistics robots designedby students have conducted preliminary application tests in the campus express delivery sorting scenarioand achieved good results.Meanwhile,the number of awards won by students in various robot competitionshas also significantly increased,which fully demonstrates the improvement of students’comprehensive qualityand practical innovation ability,and strongly proves the effectiveness and feasibility of the robot course teachingmodel based on the STEAM concept.In conclusion,the application-oriented undergraduate robot courseteaching model based on the STEAM concept constructed in this study has effectively promoted students’interdisciplinary knowledge integration,practical innovation and comprehensive quality improvement by optimizingthe curriculum system,innovating teaching methods and improving the teaching evaluation system.Itprovides new ideas and methods for the cultivation of robot professionals in application-oriented undergraduatecolleges and universities.It has significant theoretical and practical significance,and can provide reference and inspiration for the teaching reform of other related professional courses,promoting applied undergraduateeducation to better adapt to social development and industrial demands.
文摘The Outcome-Based Education(OBE)philosophy has been widely recognized as an effective ap-proach to enhance the quality of engineering education.This paper focuses on the construction of a talent cultivation system for the Robotics Engineering major based on the OBE concept.The OBE approach em-phasizes the importance of clearly defining expected learning outcomes and aligning the entire educational process to achieve these outcomes.In the context of Robotics Engineering,this means identifying the specific skills,knowledge,and competencies that graduates should possess to meet the demands of the industry and society.Firstly,the paper analyzes the current challenges and needs in thefield of Robotics Engineering.It identifies the key areas where graduates need to excel,such as programming,mechanical design,electronics,and artificial intelligence.Based on this analysis,the expected learning outcomes for the Robotics Engineering major are defined.These outcomes cover not only technical skills but also soft skills such as problem-solving,teamwork,and innovation.The paper then discusses the curriculum design that is aligned with these learning outcomes.The curriculum includes a combination of theoretical courses and practical training.The theoretical courses provide students with a solid foundation in the principles of robotics,while the practical training allows students to apply their knowledge to real-world problems.The curriculum also incorporates project-based learning,where students work on projects that simulate real-world scenarios.This approach helps students to develop their problem-solving skills and gain hands-on experience.In addition to the curriculum design,the paper also addresses the assessment methods.The assessment is outcome-based,meaning that it focuses on evaluating whether students have achieved the expected learning outcomes.This includes both formative and summative assessments,such as quizzes,exams,project evaluations,andfinal capstone projects.The as-sessment results are used to provide feedback to students and to continuously improve the teaching process.Finally,the paper presents the implementation of the talent cultivation system and evaluates its effectiveness.The results show that the OBE-based system has significantly improved the quality of education in the Robotics Engineering major.Graduates from this system are better prepared to meet the challenges of the industry and have a higher employability rate.In conclusion,the construction of a talent cultivation system for Robotics Engineering based on the OBE philosophy is an effective way to improve the quality of engineering educa-tion.By clearly defining learning outcomes and aligning the curriculum and assessment methods to achieve these outcomes,this system helps to produce well-rounded graduates who are ready to contribute to thefield.
文摘Traditional deep learning-based shadow removal methods often alter the pixels in non-shadow areas and fail to produce shadow removal results with natural boundary transitions.To solve this problem,we propose a novel multi-stage shadow removal framework based on Generative Adversarial Networks(GAN).Firstly,the multi-task-driven generator respectively generates the corresponding shadow mask and shadow mask for the input image through the shadow detection subnet and the mask generation subnet.Secondly,guided by the shadow mask and shadow mask,the full shadow module and the partial shadow module are respectively de-signed,and different types of shadows in the image are removed in stages.The sequential data from multiple sensors is used as the input of the time series generative adversarial network to generate sequence data with temporal dynamic characteristics;the data synthesized by the time series generative adversarial network is used to replace the noise input data in the gradient-penalized WGAN generative adversarial network,and the discriminator combines graph convolutional networks,long short-term memory networks and attention mech-anisms to more effectively explore the spatio-temporal correlations of multi-source heterogeneous sensing data and enhance the discriminative ability for sequential data.Then,a new combined loss function was constructed based on the least squares loss to achieve better results.Compared with the latest deep learning shadow removal methods,on the selected dataset,the balance error rate of the proposed method decreases by approximately 4.39%,the structural similarity increased by approximately 0.44%,and the pixel root mean square error de-creases by approximately 13.32%.The experimental results show that the shadow removal results obtained by this method have smoother boundary transitions.
文摘Under the backdrop of the comprehensive advancement of new engineering education,the robotics engineering discipline,which integrates multiple fields such as mechanics,electronics,control,computer science,and artificial intelligence,urgently needs to break through traditional disciplinary barriers to meet the pressing demand for interdisciplinary talents in the era of intelligent manufacturing and artificial intelligence.This paper systematically reviews the core issues existing in the current course construction of the robotics engineering major,such as a monotonous course structure,weak practical components,and insufficient interdisciplinary integration.It proposes a course system reconstruction path based on the concept of solid foundation,strong cross-disciplinary integration,emphasis on practice,and promotion of innovation.The research emphasizes that a flexible course structure of platform+module+direction should be established around industrial demands and technological frontiers,strengthening the mathematical and physical foundation and core professional courses,and expanding courses in cutting-edge areas such as artificial intelligence,big data,and cloud-edge collaboration.It also advocates for the reform of practical teaching models through project-driven,competition-education integration,and school-enterprise collaboration.Meanwhile,ideological and political education,innovation and entrepreneurship education,and engineering ethics should be organically integrated into the entire course process to form a trinity talent cultivation pattern of value shaping,knowledge imparting,and ability development.By constructing a multi-level,three-dimensional,and open course system for the robotics engineering major,it can provide strong support for the cultivation of high-level applied and innovative talents under the new engineering background,and is of great significance for promoting the connotative development of higher education and the high-quality development of industries.
文摘Artificial intelligence(AI)is profoundly reshaping the practical logic and theoretical boundaries of bel canto education.This study takes“technical path-aesthetic challenge-paradigm reconstruction”as the analytical framework to systematically explore the innovative mechanism and practical path of AI-empowered vocal education.At the technical level,an acoustic parameter modeling system is constructed based on deep learning.Through MEL spectrum analysis,forresonance tracking and breath dynamics modeling,the traditional singing assessment that relies on auditory experience is transformed into quantifiable and interpretable acoustic indicators(such as harmonic energy ratio,vowel resonance stability,etc.),achieving precise diagnosis and real-time intervention of vocal defects.Experimental data shows that this system reduces the learners’pitch error rate by 42%and increases their breath control efficiency by 35%.In the dimension of personalized training,an adaptive teaching engine based on reinforcement learning was developed to dynamically generate a three-dimensional matching scheme of“voice features-track difficulty-training intensity”,significantly reducing the time for beginners to master core vocal techniques(by an average of 28%).However,the conflict between AI quantitative indicators and the traditional aesthetic standards of vocal music has become prominent:the algorithm’s preference for standardized formuster distribution may suppress the singer’s unique timbre personality and artistic expression tension.To this end,a“dual-track evaluation model”is proposed:taking acoustic parameters as the technical benchmark and emotional expression and artistic appeal as the aesthetic benchmark,and achieving a dynamic balance between the two through expert annotation and group consensus algorithms.The research further reconstructs the teaching paradigm,advocating that AI tools be positioned as“aesthetic collaborators”rather than substitutes,and builds a new type of teacher-student relationship of“human-machine co-teaching-co-evaluation-co-creation”.This research provides a solution that is both technically feasible and aesthetically reasonable for the digital transformation of bel canto education,revealing the underlying logic of the deep integration of art and technology in the AI era.
文摘Hierarchical clustering is a widely used technique for data grouping in various fields,renowned for its ability to create a dendrogram that provides insights into data structure.However,traditional hierarchical methods often suffer from challenges such as computational inefficiency,sensitivity to noise,and difficulty in determining the optimal number of clusters.This paper introduces an improved hierarchical clustering algorithm that incorporates advanced distance metrics,the use of cluster representatives,and a robust agglomeration strategy designed to address these challenges.The proposed algorithm incorporates a mechanism to dynamically select the distance metric based on the underlying data characteristics.Unlike traditional methods that rely solely on a fixed distance metric,our approach determines the most suitable metric for the given dataset.For instance,a method like Cosine similarity could be employed for high-dimensional data or text data where orientation matters,while Euclidean distance can be effective for low-dimensional,continuous data.Our proposed method is evaluated through comprehensive experiments on synthetic and real-world datasets,showcasing significant enhancements in clustering performance,adaptability to noise,and computational efficiency.The findings indicate that the proposed algorithm outperforms traditional hierarchical methods,demonstrating its potential for broader applications across various domains.
文摘Although Chinese-English machine translation is a resource-rich language pair,the problem of data sparsity still exists.For example,for the translation of some specific domains or low-frequency words,the amount of parallel corpus is limited,which makes it difficult to improve the translation quality of the model in these scenarios.Neural machine translation models usually need a large amount of alignment data to train,otherwise they are prone to over-fitting.Neural machine translation models are slow to train and decode,especially when dealing with long sentences or complex structures,which limits their efficiency in real-time application scenarios.Chinese-English machine translation can help people overcome the language barrier and promote the communication and cooperation between people with different language backgrounds,which is of great significance for international business,academic exchanges and cultural exchanges.This paper proposes a Chinese-English neural machine translation model based on generative adversarial network.This model applies the generative adversarial network to neural machine translation,and further optimizes the adversarial learning based neural machine translation model by improving the monotone decoding sequence from left to right or from right to left in the original machine translation model.At the same time,unlike previous generative adversarial networks,neural machine translation models are actually a sequence of discrete symbols that map source language sentences to target language sentences,both in discontinuous spaces.In this case,the generative adversarial network fails to transmit the gradient properly,causing the generator to lose its update direction.By introducing the strategy gradient algorithm in reinforcement learning,the generator optimization problem in adversarial learning is solved,and the translation performance of the model is improved.Finally,experiments on public data sets show that the proposed model can effectively improve translation quality compared with other advanced models.
基金supported by the School-level Science Research Project of Weifang Vocational College under Grant No.WZZX202431.
文摘In the new era,vocational college has been an important component of the education system in China,responsible for the mission of cultivating high-quality technical talents.The construction of one-stop student communities has become an important way to promote the modernization of vocational education governance.With the deepening of education reform,vocational colleges are facing new challenges in the con-struction of student communities.The contradiction between the expansion of the scale of vocational college students and the diversification of demand is prominent,and the traditional management model has shortcom-ings such as scattered resources and insufficient collaboration.How to explore a one-stop student community construction path that is in line with the characteristics of vocational colleges has become an urgent issue to be solved.Nowadays,the development of artificial intelligence(AI)technology is flourishing,and deeply changing the face of higher education.The rapid development of AI technology provides strong support for the construction of student communities in vocational colleges.AI intelligent tools,big data information col-lection and other technologies,can effectively help colleges continuously optimize their management models,greatly improve service efficiency and quality,and enhance student satisfaction.Hence,based on the AI tech-nology,building an equal,inclusive,and interactive student community network,responding to students’needs for digital learning resources,platform tools,service support,and improving the accuracy and satisfaction of education management,are of great significance for the vocational education reform in China.This article overviews pervious achievements of AI-enabled student community construction in colleges,to provide ac-tionable solutions for the paradigm shift of the student service system in vocational colleges at the practical level.
文摘In the pedestrian detection scenario,the detection algorithm usually misses obscured and distant fuzzy pedestrians,and at the same time cannot take into account the detection accuracy and speed.In this paper,we propose a modified YOLOv5 model for pedestrian detection.Firstly,the backbone network uses the SPD-GCONV module constructed by the combination of SPD(Space-to-Depth)module and Ghost convolution for down-sampling to reduce the loss of fine-grained feature information.Secondly,the multi-scale detection ability of the model is enhanced by adding a small size detection layer.Then,the original CIoU loss function is replaced by α-EloU loss function to improve the accuracy of pedestrian target location.According to the experiments on WiderPerson data set,the average detection accuracy is improved by 2%compared with other pedestrian detection algorithms on the premise of ensuring the detection speed.Experimental results show that the improved algorithm can significantly improve the detection performance.
文摘Fourth party logistics(4PL)supply chain network design problem considering the behavior of customer service satisfaction becomes more important in supply chain network design problem.It is NP-hard problem.Improved differential evolution(IDE)algorithm is applied to solve this problem in this paper.In order to handle the infeasible solution efficiently,a two-stage algorithm,the minimum cost flow incorporated with macro-micro adjustment embedded improved differential evolution(MMAMFC-IDE),is designed.The results of numerical experiments show that MMAMFC-IDE can get relatively better results compared with the minimum cost flow embedded improved differential evolution(MFC-IDE).
基金supported by The Ministry of Education’s Industry-University Cooperation and Collaborative Education Project,“Practical Exploration of AI Vision Project-based Training for Artificial Intelligence Teachers in Robotics Engineering”,Project Number:2506205737.
文摘Industry-education integration communities are redefining how advanced robotics courses are designed,delivered and validated.This paper reports a two-year design-based study that reconceptualised the undergraduate module Robot Vision Perception and Detection from a supply-chain-oriented alliance of three universities,two robotics manufacturers and one logistics giant.Adopting a community-of-practice lens,we replaced the conventional lectureClab sequence with a challenge-driven co-creation loop in which corporate engineers release real-time production-line vision defects as curricular tasks,faculty scaffold theoretical principles,and students iterate solutions on the factory floor using identical hardware and data streams.Mixedmethods evaluation with 142 students and 18 industry mentors shows significant gains:(1)learning performance increased by 0.82 standard deviations;(2)student creative self-efficacy and systems-thinking improved 34%and 29%respectively;(3)average defect detection recall of student models rose from 72%to 93%,with 8 prototypes transferred to the partner lines;(4)facultyCindustry co-publications and patent disclosures tripled.Qualitative trace data reveal that boundary objects(annotated datasets,Dockerized algorithms and shared Kanban boards)legitimately brokered epistemic differences between academia and industry.The study contributes an empirically grounded frameworkłCIE-CDL(Community-Integrated Education via Challenge-Driven Loops)łthat embeds authentic socio-technical complexity into robotics curricula while simultaneously generating measurable value for industrial partners.Implications for scalable,sustainable industryCeducation symbiosis are discussed.
文摘The rapid development of artificial intelligence technology has provided an opportunity to reshape the teaching ecosystem in open education.This article focuses on the concept of“adaptive learning”,in the context of the artificial intelligence era,and explores the systematic reform of open education teaching models.The researchfirst constructed an integrated learning framework that combines cognitive diagnosis,dynamic paths,resource push,immediate feedback,and emotional support.Through data-driven and teacher experience collaboration,it realizes large-scale personalized teaching.Secondly,based on the teaching practice of public courses in multiple universities,the article collected and analyzed the entire process behavior data of learners,used deep models to dynamically optimize teaching strategies,and established an interpretable and iterative teaching loop.On this basis,the research focuses on educational equity and the mechanism of human-computer collaboration,ensuring that while technology is empowered,the dominant position of teachers and the warmth of the learning community are maintained.Through qualitative interviews and teaching observations,the article found that adaptive learning significantly enhanced the initiative,satisfaction,and knowledge transfer ability of learners,forming a new classroom culture that integrates online and offline elements and reshapes the roles of teachers and students.The research conclusion states that in the open education teaching reform of the artificial intelligence era,it should be driven by data intelligence,centered on learners,and based on educational equity,promoting the transformation from“standardized supply”to“precise services”,providing replicable models and sustainable paths for building a lifelong learning society.
基金supported by the Project of the Research Center for the Theory System of Socialism with Chinese Characteristics in Shenyang Normal University,2023.Project number:ZTSYB2023012.Project name:A Study of Xi Jinping Thought on Socialism with Chinese Characteristics in the Course of Foreign Affairs Translation.
文摘In recent years,neural machine translation has rapidly developed and replaced traditional machine translation,becoming the mainstream paradigm in the field of machine translation.Machine translation can reduce translation costs and improve translation efficiency,bring good news to cultural exchanges and international cooperation,and help national development.However,neural machine translation is highly dependent on large-scale high-quality parallel corpus,and there are problems such as uneven quality and sparse data,so it is imperative to study and explore neural machine translation.The purpose of this paper is to construct pseudo-parallel corpus using data enhancement technology,improve the diversity of Chinese and English materials,and then optimize the translation model to improve the translation effect of the model.Based on BERT pre-training technology,this paper first analyzes the limitations of the traditional Transformer model,and then puts forward two directions for model optimization.On the one hand,in the data preprocessing stage,multi-granularity word segmentation technology is used for word segmentation to help Chinese-English neural machine translation model better understand the text.On the other hand,in the pre-training stage,this paper adopts the strategy of deep integration of BERT dynamic word embedding and original word embedding.On the basis of the original Transformer,a fusion module is added,through which the original word embeddings and BERT dynamic word embeddings are simple linear splicing,and then fed into the encoder.The attention mechanism is used for deep integration and better word vector representation,enabling the Transformer model to take full advantage of the external semantic information introduced by BERT.Finally,the feasibility and effectiveness of the Transformer architecture adopted in this paper are verified by the comparison experiment between RNN and Transformer model.Through the ablation experiments of different word vector representation and different stages using BERT pre-training technology,the effectiveness of BERT dynamic word embedding and deep fusion of word embedding and the rationality of using pre-training technology only in the encoder stage are verified.
文摘This paper proposes a novel image enhancement based on guidedfiltering and modified Retinex method to address the problems of poor visual effect,unclear images and incomplete display of details in low-light images.Low-illumination images are processed using guidedfiltering to adaptively enhance the luminance component.This method consists of two parts:the image decomposition network and the image en-hancement network.The image decomposition network is responsible for decomposing the original image into illumination components and reflection components.The image enhancement network optimizes parameters and performsγcorrection through the natural image quality evaluator(NIQE),adjusting the brightness and contrast of the illumination component,and then re-fuses it with the reflection component to improve the over-all image quality.Finally,image quality enhancement experiments are conducted using the images in the SOTS standard test set and six actual scene images.They are compared and analyzed with other advanced methods.The experimental results show that the proposed method is superior to other methods in both subjective visual effects and objective evaluation indicators,thus fully verifying the effectiveness and feasibility of the proposed method.
文摘Due to the large scale and high dimension of teaching data,the using of traditional clustering algorithms has problems such as high computational complexity and low accuracy.Therefore,this paper proposes a weighted block sparse subspace clustering algorithm based on information entropy.The introduction of information entropy weight and block diagonal constraints can obtain the prior probability that two pixels belong to the same category before the simulation experiment,thereby positively intervening that the solutions solved by the model tend to be the optimal approximate solutions of the block diagonal structure.It can enable the model to obtain the performance against noise and outliers,and thereby improving the discriminative ability of the model classification.The experimental results show that the average probability Rand index of the proposed method is 0.86,which is higher than that of other algorithms.The average information change index of the proposed method is 1.55,which is lower than that of other algorithms,proving its strong robustness.On different datasets,the misclassification rates of the design method are 1.2%and 0.9%respectively,which proves that its classification accuracy is relatively high.The proposed method has high reliability in processing highdimensional teaching data.It can play an important role in the field of educational data analysis and provide strong support for intelligent teaching.
基金supported by the“Research on language adaptive model based on Deep learning in futureoriented teaching scenarios”,project number:25B413010.
文摘Deep learning is more and more widely used in natural language processing.Compared with the traditional n-gram statistical language model,Recurrent neural network(RNN)modeling technology has shown great advantages in language modeling,and has been gradually applied in speech recognition,machine translation and other fields.However,at present,the training of RNN language models is mostly offline.For different speech recognition tasks,there are language differences between training corpus and recognition tasks,which affects the recognition rate of speech recognition systems.While using RNN modeling technology to train the Chinese language model,an online RNN model self-adaption algorithm is proposed,which takes the preliminary recognition results of speech signals as corpus to continue training the model,so that the adaptive RNN model can get the maximum match with the recognition task.The experimental results show that the adaptive model effectively reduces the language difference between the language model and the recognition task,and the recognition rate of the system is further improved after the Chinese word confusion network is re-scored,which has been verified in the actual Chinese speech recognition system.
文摘This study designs a curriculum system that deeply integrates industry and education for robotics engineering majors in applied universities.Guided by the CDIO philosophy and the concept of industryCeducation collaboration,a three-level curriculum framework consisting of foundational courses,professional cores,and integrated practical modules is constructed.Industry technologies and standards are introduced into course content,teaching methods,and evaluation systems through co-developed courses,shared laboratories,and joint faculty teams.The system emphasizes continuous engagement with real-world engineering projects,fostering students’competencies in design,programming,integration,and innovation.The proposed model provides a flexible and sustainable approach for cultivating application-oriented robotics talents that meet the evolving needs of intelligent manufacturing.
基金supported by the Henan Provincial Scientific and Technological Research Project(No.252102210005,222102310222)the Training Program for Young Backbone Teachers in Higher Education Institutions of Henan Province(No.2025GGJS149).
文摘This paper proposes a novel arrhythmia classification method combining Convolutional Neural Networks(CNN),Attention mechanisms,and Bidirectional Transformers(BiTransformer).The method aims to improve the accuracy and robustness of arrhythmia detection in electrocardiogram(ECG)signals.Initially,the CNN module extracts local spatial features from raw ECG data,effectively capturing the morphological characteristics of different arrhythmia types.Subsequently,the Attention mechanism is applied to weigh the importance of different segments in the ECG signal,allowing the model to focus on critical features that are most indicative of arrhythmia.Finally,the BiTransformer module processes the feature sequences bidirectionally,capturing both forward and backward dependencies in the ECG signal.This comprehensive approach enables the model to integrate local and global information,enhancing its ability to classify various arrhythmias accurately.Experiments conducted on the MIT-BIH Arrhythmia Database demonstrate that the proposed method achieves state-of-the-art performance,with a significant improvement in classification accuracy compared to traditional methods.The results highlight the effectiveness of combining CNN,Attention,and BiTransformer for arrhythmia classification,offering a promising direction for automated ECG analysis and clinical applications.
文摘This study is based on the theory of human engineering psychology and explores the mechanism by which cool and warm color posters affect the emotions of college students.Through the combination of exper-imental methods and scale measurement,60 college students were selected as participants.Cool color(blue,green)and warm color(red,yellow)posters were randomly presented to them,and their emotional responses were collected.The SAM scale(Self-Assessment Manikin)was used to assess the degree of pleasure,arousal,and dominance,and skin conductance response(GSR)was used to obtain physiological indicators to enhance the objectivity of the study.The results showed that warm color posters significantly enhanced the arousal and pleasure of college students,especially under the red stimulus,with the highest level of emotional activation.While cool color posters significantly reduced arousal and brought a more calm and relaxed emotional expe-rience.The blue tone was the most effective in alleviating anxiety.Gender and professional background have a moderating effect on color perception.Women are more sensitive to the emotional responses to cool and warm colors,and students majoring in art design have a higher recognition and emotional response intensity of color differences than non-art majors.This study verified the applicability of color emotional effects in the col-lege student population and expanded the application boundaries of human engineering psychology in visual communication design.The research suggests that in college mental health publicity and campus environment design,scientific application of color psychology principles should be adopted,and appropriate colors should be selected according to different emotional regulation goals to improve the accuracy and effectiveness of emo-tional intervention.In the future,neuroscientific methods such as electroencephalogram(EEG)can be further introduced to deeply reveal the neural basis of the color emotional mechanism.
