In the very beginning,the Computer Laboratory of the University of Cambridge was founded to provide computing service for different disciplines across the university.As computer science developed as a discipline in it...In the very beginning,the Computer Laboratory of the University of Cambridge was founded to provide computing service for different disciplines across the university.As computer science developed as a discipline in its own right,boundaries necessarily arose between it and other disciplines,in a way that is now often detrimental to progress.Therefore,it is necessary to reinvigorate the relationship between computer science and other academic disciplines and celebrate exploration and creativity in research.To do this,the structures of the academic department have to act as supporting scaffolding rather than barriers.Some examples are given that show the efforts being made at the University of Cambridge to approach this problem.展开更多
Since its inaugural issue in 1986,the Journal of Computer Science and Technology(JCST)has been the premier English journal of China Computer Federation(CCF),serving international readers and authors by disseminating s...Since its inaugural issue in 1986,the Journal of Computer Science and Technology(JCST)has been the premier English journal of China Computer Federation(CCF),serving international readers and authors by disseminating scholarly and technical papers under a rigorous review process.展开更多
With the rapid development of China’s economy,the demand for high-end talents in the field of engineering technology is becoming increasingly prominent.Engineering doctors,as an important force in this field,have a d...With the rapid development of China’s economy,the demand for high-end talents in the field of engineering technology is becoming increasingly prominent.Engineering doctors,as an important force in this field,have a direct impact on the progress of national technological innovation and the upgrading of industrial structure.Currently,there are still some issues in the university-enterprise collaboration for engineering doctor training in China,such as unclear cooperation mechanisms and responsibility divisions,insufficient corporate participation and enthusiasm,and imperfect evaluation and feedback mechanisms.This paper aims to explore the university-enterprise collaborative training path of engineering doctors from the perspective of deep integration,analyzing multiple dimensions including training objectives,curriculum design,practical sessions,mentor teams,and evaluation systems,in order to provide reference for the reform and practice of engineering doctor training in China.展开更多
This study explores the feasibility of constructing an intelligent educational evaluation system based on the CIPP model and artificial intelligence technology in the context of new engineering disciplines.By integrat...This study explores the feasibility of constructing an intelligent educational evaluation system based on the CIPP model and artificial intelligence technology in the context of new engineering disciplines.By integrating the CIPP model with AI technology,a novel intelligent educational evaluation system was designed.Through experimental validation and case studies,the system demonstrated significant effectiveness in improving teaching quality,facilitating personalized student development,and optimizing educational resource allocation.Additionally,the study predicts potential changes this system could bring to the education industry and proposes relevant policy recommendations.Although the current research has limitations,with technological advancements in the future,this system is expected to provide stronger support for innovations in engineering education models.展开更多
Plant diseases are a major threat that can severely impact the production of agriculture and forestry.This can lead to the disruption of ecosystem functions and health.With its ability to capture continuous narrow-ban...Plant diseases are a major threat that can severely impact the production of agriculture and forestry.This can lead to the disruption of ecosystem functions and health.With its ability to capture continuous narrow-band spectra,hyperspectral technology has become a crucial tool to monitor crop diseases using remote sensing.However,existing continuous wavelet analysis(CWA)methods suffer from feature redundancy issues,while the continuous wavelet projection algorithm(CWPA),an optimization approach for feature selection,has not been fully validated to monitor plant diseases.This study utilized rice bacterial leaf blight(BLB)as an example by evaluating the performance of four wavelet basis functions-Gaussian2,Mexican hat,Meyer,andMorlet-within theCWAandCWPAframeworks.Additionally,the classification models were constructed using the k-nearest neighbors(KNN),randomforest(RF),and Naïve Bayes(NB)algorithms.The results showed the following:(1)Compared to traditional CWA,CWPA significantly reduced the number of required features.Under the CWPA framework,almost all the model combinations achieved maximum classification accuracy with only one feature.In contrast,the CWA framework required three to seven features.(2)Thechoice of wavelet basis functions markedly affected the performance of themodel.Of the four functions tested,the Meyer wavelet demonstrated the best overall performance in both the CWPA and CWA frameworks.(3)Under theCWPAframework,theMeyer-KNNandMeyer-NBcombinations achieved the highest overall accuracy of 93.75%using just one feature.In contrast,under the CWA framework,the CWA-RF combination achieved comparable accuracy(93.75%)but required six features.This study verified the technical advantages of CWPA for monitoring crop diseases,identified an optimal wavelet basis function selection scheme,and provided reliable technical support to precisely monitor BLB in rice(Oryza sativa).Moreover,the proposed methodological framework offers a scalable approach for the early diagnosis and assessment of plant stress,which can contribute to improved accuracy and timeliness when plant stress is monitored.展开更多
Rockfall kinematic characteristics exhibit significant randomness and are influenced by factors such as rock mass properties,slope morphology,impact angle,and slope materials.Accurately determining the key parameters ...Rockfall kinematic characteristics exhibit significant randomness and are influenced by factors such as rock mass properties,slope morphology,impact angle,and slope materials.Accurately determining the key parameters of rockfall movement is critical for understanding motion patterns and effectively preventing and controlling rockfall hazards.In this study,a monitoring system consisting of selfdeveloped inertial navigation equipment,high-speed cameras,and an unmanned aerial vehicle was used to conduct onsite motion tests involving four differently shaped rock specimens on three types of slopes(bedrock,detritus,and clast bedding).The selfdeveloped inertial navigation system integrated a highdynamic-range accelerometer(±400 g)and a shockresistant gyroscope(±4000°/s),capable of robustly collecting data during the test.The data collected from these tests were processed to extract key kinematic parameters such as velocity,trajectory,restitution coefficients,and friction coefficients.The test results demonstrated that the inertial navigation system accurately recorded the acceleration and angular velocity of the rocks during motion,with these measurements closely aligning with the field data.The normal and tangential restitution coefficients were found to be influenced primarily by the slope material and impact angle,with higher normal restitution coefficients observed for low-angle impacts.The normal restitution coefficients ranged from 0.35 to 0.86,whereas the tangential restitution coefficients ranged from 0.46 to 0.91,depending on the slope materials.Additionally,the sliding friction coefficient was calculated to be between 0.66 and 0.78,whereas the rolling friction coefficient for the slab-shaped specimen was determined to be 0.53.These findings provide valuable data for improving the accuracy of rockfall trajectory predictions and the design of protective structures.展开更多
To enhance speech emotion recognition capability,this study constructs a speech emotion recognition model integrating the adaptive acoustic mixup(AAM)and improved coordinate and shuffle attention(ICASA)methods.The AAM...To enhance speech emotion recognition capability,this study constructs a speech emotion recognition model integrating the adaptive acoustic mixup(AAM)and improved coordinate and shuffle attention(ICASA)methods.The AAM method optimizes data augmentation by combining a sample selection strategy and dynamic interpolation coefficients,thus enabling information fusion of speech data with different emotions at the acoustic level.The ICASA method enhances feature extraction capability through dynamic fusion of the improved coordinate attention(ICA)and shuffle attention(SA)techniques.The ICA technique reduces computational overhead by employing depth-separable convolution and an h-swish activation function and captures long-range dependencies of multi-scale time-frequency features using the attention weights.The SA technique promotes feature interaction through channel shuffling,which helps the model learn richer and more discriminative emotional features.Experimental results demonstrate that,compared to the baseline model,the proposed model improves the weighted accuracy by 5.42%and 4.54%,and the unweighted accuracy by 3.37%and 3.85%on the IEMOCAP and RAVDESS datasets,respectively.These improvements were confirmed to be statistically significant by independent samples t-tests,further supporting the practical reliability and applicability of the proposed model in real-world emotion-aware speech systems.展开更多
In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and rec...In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and recognition accuracy on resource-constrained devices such as mobile terminals remains a major challenge.To address this,we propose a novel time-frequency dual-branch parallel residual network,which integrates a Dual-Branch Broadcast Residual module and a Time-Frequency Coordinate Attention module.The time-domain and frequency-domain branches are designed in parallel to independently extract temporal and spectral features,effectively avoiding the potential information loss caused by serial stacking,while enhancing information flow and multi-scale feature fusion.In terms of training strategy,a curriculum learning approach is introduced to progressively improve model robustness fromeasy to difficult tasks.Experimental results demonstrate that the proposed method consistently outperforms existing lightweight models under various signal-to-noise ratio(SNR)conditions,achieving superior far-field recognition performance on the Google Speech Commands V2 dataset.Notably,the model maintains stable performance even in low-SNR environments such as–10 dB,and generalizes well to unseen SNR conditions during training,validating its robustness to novel noise scenarios.Furthermore,the proposed model exhibits significantly fewer parameters,making it highly suitable for deployment on resource-limited devices.Overall,the model achieves a favorable balance between performance and parameter efficiency,demonstrating strong potential for practical applications.展开更多
With the rapid development of smart manufacturing,intelligent safety monitoring in industrial workshops has become increasingly important.To address the challenges of complex backgrounds,target scale variation,and exc...With the rapid development of smart manufacturing,intelligent safety monitoring in industrial workshops has become increasingly important.To address the challenges of complex backgrounds,target scale variation,and excessive model parameters in worker violation detection,this study proposes ADCP-YOLO,an enhanced lightweight model based on YOLOv8.Here,“ADCP”represents four key improvements:Alterable Kernel Convolution(AKConv),Dilated-Wise Residual(DWR)module,Channel Reconstruction Global Attention Mechanism(CRGAM),and Powerful-IoU loss.These components collaboratively enhance feature extraction,multi-scale perception,and localization accuracy while effectively reducing model complexity and computational cost.Experimental results show that ADCP-YOLO achieves a mAP of 90.6%,surpassing YOLOv8 by 3.0%with a 6.6%reduction in parameters.These findings demonstrate that ADCP-YOLO successfully balances accuracy and efficiency,offering a practical solution for intelligent safety monitoring in smart factory workshops.展开更多
The colorectal cancer is one of the most common and lethal cancers,and colorectal polyps,as precancerous lesions,can lead to diagnostic oversight or misdiagnosis due to their varied shapes and sizes,thereby promoting ...The colorectal cancer is one of the most common and lethal cancers,and colorectal polyps,as precancerous lesions,can lead to diagnostic oversight or misdiagnosis due to their varied shapes and sizes,thereby promoting the irreversible progression of colorectal cancer.We propose a YOLO based model and name it EF-YOLO.It incorporates transformer to extract contextual information about the colorectal polyps.Simultaneously,leveraging the morphological characteristics of colorectal polyps,we design a brand-new module,namely advanced multi-scale aggregation(AMSA),to replace the traditional multi-scale module.The backbone adopts deformable convolutional network-maxpool(DCN-MP)to enhance feature extraction while adaptively sampling points to better match the shapes of colorectal polyps.By combining coordinate attention(CA),this model maximizes the use of positional and channel information,more effectively extracting features of colorectal polyps,directing the model’s attention toward the colorectal polyp region.EF-YOLO has made advancement on the merged Kvasir-SEG and CVC-ClinicDB dataset.Compared to the original model,the mean average precision(mAP)of EF-YOLO increases and reaches 96.60%,meeting automated colorectal polyp detection requirements.展开更多
In non-independent and identically distributed(non-IID)data environments,model performance often degrades significantly.To address this issue,two improvement methods are proposed:FedReg and FedReg^(*).FedReg is a meth...In non-independent and identically distributed(non-IID)data environments,model performance often degrades significantly.To address this issue,two improvement methods are proposed:FedReg and FedReg^(*).FedReg is a method based on hybrid regularization aimed at enhancing federated learning in non-IID scenarios.It introduces hybrid regularization to replace traditional L2 regularization,combining the advantages of L1 and L2 regularization to enable feature selection while preventing overfitting.This method better adapts to the diverse data distributions of different clients,improving the overall model performance.FedReg^(*)combines hybrid regularization with weighted model aggregation.In addition to the benefits of hybrid regularization,FedReg^(*)applies a weighted averaging method in the model aggregation process,calculating weights based on the cosine similarity between each client gradient and the global gradient to more reasonably distribute client contributions.By considering variations in data quality and quantity among clients,FedReg^(*)highlights the importance of key clients and enhances the model’s generalization performance.These improvement methods enhance model accuracy and communication efficiency.展开更多
This study presents a teaching reform for the Object-oriented Software Construction(OOSC)course by integrating outcome-based education(OBE)and the BOPPPS(bridge-In,objectives,pre-assessment,participatory learning,post...This study presents a teaching reform for the Object-oriented Software Construction(OOSC)course by integrating outcome-based education(OBE)and the BOPPPS(bridge-In,objectives,pre-assessment,participatory learning,post-assessment,summary)instructional model.The reform addresses the gap between syntax-based programming instruction and the need for higher-level skills in abstraction,modularity,and software architecture.The course is anchored in a semester-long,project-based learning platform centered on a Java-based Aircraft Battle Game,progressing through six iterative experiments.Each experiment targets specific competencies within the structured BOPPPS teaching cycle and is aligned with specific OBE learning outcomes.A case study on the Factory Pattern illustrates how the BOPPPS model fosters conceptual understanding and practical application.Evaluation results from the 2023 and 2024 spring semesters show improved outcomes:Project completion rose from 87%to 95%,37%of students implemented innovative features,and average final grades increased by 7%.The results affirm that the OBE+BOPPPS integration strengthens engagement,deepens understanding,and equips students with real-world software development competencies.展开更多
Small object detection has been a focus of attention since the emergence of deep learning-based object detection.Although classical object detection frameworks have made significant contributions to the development of...Small object detection has been a focus of attention since the emergence of deep learning-based object detection.Although classical object detection frameworks have made significant contributions to the development of object detection,there are still many issues to be resolved in detecting small objects due to the inherent complexity and diversity of real-world visual scenes.In particular,the YOLO(You Only Look Once)series of detection models,renowned for their real-time performance,have undergone numerous adaptations aimed at improving the detection of small targets.In this survey,we summarize the state-of-the-art YOLO-based small object detection methods.This review presents a systematic categorization of YOLO-based approaches for small-object detection,organized into four methodological avenues,namely attention-based feature enhancement,detection-head optimization,loss function,and multi-scale feature fusion strategies.We then examine the principal challenges addressed by each category.Finally,we analyze the performance of thesemethods on public benchmarks and,by comparing current approaches,identify limitations and outline directions for future research.展开更多
Traffic sign detection is a critical component of driving systems.Single-stage network-based traffic sign detection algorithms,renowned for their fast detection speeds and high accuracy,have become the dominant approa...Traffic sign detection is a critical component of driving systems.Single-stage network-based traffic sign detection algorithms,renowned for their fast detection speeds and high accuracy,have become the dominant approach in current practices.However,in complex and dynamic traffic scenes,particularly with smaller traffic sign objects,challenges such as missed and false detections can lead to reduced overall detection accuracy.To address this issue,this paper proposes a detection algorithm that integrates edge and shape information.Recognizing that traffic signs have specific shapes and distinct edge contours,this paper introduces an edge feature extraction branch within the backbone network,enabling adaptive fusion with features of the same hierarchical level.Additionally,a shape prior convolution module is designed to replaces the first two convolutional modules of the backbone network,aimed at enhancing the model's perception ability for specific shape objects and reducing its sensitivity to background noise.The algorithm was evaluated on the CCTSDB and TT100k datasets,and compared to YOLOv8s,the mAP50 values increased by 3.0%and 10.4%,respectively,demonstrating the effectiveness of the proposed method in improving the accuracy of traffic sign detection.展开更多
War rehearsals have become increasingly important in national security due to the growing complexity of international affairs.However,traditional rehearsal methods,such as military chess simulations,are inefficient an...War rehearsals have become increasingly important in national security due to the growing complexity of international affairs.However,traditional rehearsal methods,such as military chess simulations,are inefficient and inflexible,with particularly pronounced limitations in command and decision-making.The overwhelming volume of information and high decision complexity hinder the realization of autonomous and agile command and control.To address this challenge,an intelligent warfare simulation framework named Command-Agent is proposed,which deeply integrates large language models(LLMs)with digital twin battlefields.By constructing a highly realistic battlefield environment through real-time simulation and multi-source data fusion,the natural language interaction capabilities of LLMs are leveraged to lower the command threshold and to enable autonomous command through the Observe-Orient-Decide-Act(OODA)feedback loop.Within the Command-Agent framework,a multimodel collaborative architecture is further adopted to decouple the decision-generation and command-execution functions of LLMs.By combining specialized models such as Deep Seek-R1 and MCTool,the limitations of single-model capabilities are overcome.MCTool is a lightweight execution model fine-tuned for military Function Calling tasks.The framework also introduces a Vector Knowledge Base to mitigate hallucinations commonly exhibited by LLMs.Experimental results demonstrate that Command-Agent not only enables natural language-driven simulation and control but also deeply understands commander intent.Leveraging the multi-model collaborative architecture,during red-blue UAV confrontations involving 2 to 8 UAVs,the integrated score is improved by an average of 41.8%compared to the single-agent system(MCTool),accompanied by a 161.8%optimization in the battle loss ratio.Furthermore,when compared with multi-agent systems lacking the knowledge base,the inclusion of the Vector Knowledge Base further improves overall performance by 16.8%.In comparison with the general model(Qwen2.5-7B),the fine-tuned MCTool leads by 5%in execution efficiency.Therefore,the proposed Command-Agent introduces a novel perspective to the military command system and offers a feasible solution for intelligent battlefield decision-making.展开更多
Amid the increasing demand for data sharing,the need for flexible,secure,and auditable access control mechanisms has garnered significant attention in the academic community.However,blockchain-based ciphertextpolicy a...Amid the increasing demand for data sharing,the need for flexible,secure,and auditable access control mechanisms has garnered significant attention in the academic community.However,blockchain-based ciphertextpolicy attribute-based encryption(CP-ABE)schemes still face cumbersome ciphertext re-encryption and insufficient oversight when handling dynamic attribute changes and cross-chain collaboration.To address these issues,we propose a dynamic permission attribute-encryption scheme for multi-chain collaboration.This scheme incorporates a multiauthority architecture for distributed attribute management and integrates an attribute revocation and granting mechanism that eliminates the need for ciphertext re-encryption,effectively reducing both computational and communication overhead.It leverages the InterPlanetary File System(IPFS)for off-chain data storage and constructs a cross-chain regulatory framework—comprising a Hyperledger Fabric business chain and a FISCO BCOS regulatory chain—to record changes in decryption privileges and access behaviors in an auditable manner.Security analysis shows selective indistinguishability under chosen-plaintext attack(sIND-CPA)security under the decisional q-Parallel Bilinear Diffie-Hellman Exponent Assumption(q-PBDHE).In the performance and experimental evaluations,we compared the proposed scheme with several advanced schemes.The results show that,while preserving security,the proposed scheme achieves higher encryption/decryption efficiency and lower storage overhead for ciphertexts and keys.展开更多
Workflow scheduling is critical for efficient cloud resource management.This paper proposes Tunicate Swarm-Highest Response Ratio Next,a novel scheduler that synergistically combines the Tunicate Swarm Algorithm with ...Workflow scheduling is critical for efficient cloud resource management.This paper proposes Tunicate Swarm-Highest Response Ratio Next,a novel scheduler that synergistically combines the Tunicate Swarm Algorithm with the Highest Response Ratio Next policy.The Tunicate Swarm Algorithm generates a cost-minimizing task-to-VM mapping scheme,while the Highest Response Ratio Next dynamically dispatches tasks in the ready queue with the highest-priority.Experimental results demonstrate that the Tunicate Swarm-Highest Response RatioNext reduces costs by up to 94.8%compared to meta-heuristic baselines.It also achieves competitive cost efficiency vs.a learning-based method while offering superior operational simplicity and efficiency,establishing it as a highly practical solution for dynamic cloud environments.展开更多
Since the advent of smart contracts,security vulnerabilities have remained a persistent challenge,compromsing both the reliability of contract execution and the overall stability of the virtual currency market.Consequ...Since the advent of smart contracts,security vulnerabilities have remained a persistent challenge,compromsing both the reliability of contract execution and the overall stability of the virtual currency market.Consequently,the academic community has devoted increasing attention to these security risks.However,conventional approaches to vulnerability detection frequently exhibit limited accuracy.To address this limitation,the present study introduces a novel vulnerability detection framework called GNNSE that integrates symbolic execution with graph neural networks(GNNs).The proposedmethod first constructs semantic graphs to comprehensively capture the control flow and data flow dependencies within smart contracts.These graphs are subsequently processed using GNNs to efficiently identify contracts with a high likelihood of vulnerabilities.For these high-risk contracts,symbolic execution is employed to perform fine-grained,path-level analysis,thereby improving overall detection precision.Experimental results on a dataset comprising 10,079 contracts demonstrate that the proposed method achieves detection precisions of 93.58% for reentrancy vulnerabilities and 92.73% for timestamp-dependent vulnerabilities.展开更多
Attribute-Based Encryption(ABE)has emerged as a fundamental access control mechanism in data sharing,enabling data owners to define flexible access policies.A critical aspect of ABE is key revocation,which plays a piv...Attribute-Based Encryption(ABE)has emerged as a fundamental access control mechanism in data sharing,enabling data owners to define flexible access policies.A critical aspect of ABE is key revocation,which plays a pivotal role in maintaining security.However,existing key revocation mechanisms face two major challenges:(1)High overhead due to ciphertext and key updates,primarily stemming from the reliance on revocation lists during attribute revocation,which increases computation and communication costs.(2)Limited universality,as many attribute revocation mechanisms are tailored to specific ABE constructions,restricting their broader applicability.To address these challenges,we propose LUAR(Lightweight and Universal Attribute Revocation),a novel revocation mechanism that leverages Intel Software Guard Extensions(SGX)while minimizing its inherent limitations.Given SGX’s constrained memory(≈90 MB in a personal computer)and susceptibility to side-channel attacks,we carefully manage its usage to reduce reliance while mitigating potential collusion risks between cloud service providers and users.To evaluate LUAR’s lightweight and universality,we integrate it with the classic BSW07 scheme,which can be seamlessly replaced with other ABE constructions.Experimental results demonstrate that LUAR enables secure attribute revocation with low computation and communication overhead.The processing time within the SGX environment remains stable at approximately 55 ms,regardless of the complexity of access policies,ensuring no additional storage or computational burden on SGX.Compared to the Hardware-based Revocable Attribute-Based Encryption(HR-ABE)scheme(IEEE S&P 2024),LUAR incurs a slightly higher computational cost within SGX;however,the overall time from initiating a data request to obtaining plaintext is shorter.As access policies grow more complex,LUAR’s advantages become increasingly evident,showcasing its superior efficiency and broader applicability.展开更多
Training large-scale deep neural networks(DNNs)is prone to software and hardware failures,with critical failures often requiring full-machine reboots that substantially prolong training.Existing checkpoint-recovery so...Training large-scale deep neural networks(DNNs)is prone to software and hardware failures,with critical failures often requiring full-machine reboots that substantially prolong training.Existing checkpoint-recovery solutions either cannot tolerate such critical failures or suffer from slow checkpointing and recovery due to constrained input/output bandwidth.In this paper,we propose FastCheck,a checkpoint-recovery framework that accelerates checkpointing and recovery through parallel transmission and tailored compression.First,FastCheck partitions checkpoints into shards and leverages multiple nodes for parallel checkpointing and recovery.Second,it further reduces checkpoint size and overhead with delta compression for weights and index compression for momentum.Third,FastCheck employs lightweight and consistent health status maintenance that accurately tracks node health,preventing checkpoint transmission to failed nodes.We implement FastCheck in PyTorch and evaluate it on multiple DNN models against two baselines.Experimental results show that FastCheck reduces the checkpointing time by up to 78.42%and the recovery time by up to 77.41%,while consistently improving efficiency across different training stages.展开更多
文摘In the very beginning,the Computer Laboratory of the University of Cambridge was founded to provide computing service for different disciplines across the university.As computer science developed as a discipline in its own right,boundaries necessarily arose between it and other disciplines,in a way that is now often detrimental to progress.Therefore,it is necessary to reinvigorate the relationship between computer science and other academic disciplines and celebrate exploration and creativity in research.To do this,the structures of the academic department have to act as supporting scaffolding rather than barriers.Some examples are given that show the efforts being made at the University of Cambridge to approach this problem.
文摘Since its inaugural issue in 1986,the Journal of Computer Science and Technology(JCST)has been the premier English journal of China Computer Federation(CCF),serving international readers and authors by disseminating scholarly and technical papers under a rigorous review process.
基金supported in part by the 2023 College Student Innovation and Entrepreneurship Training 430 Program Project of China(Grant No.202310699159)2024 Graduate Education Comprehensive Reform,Development and Innovation Project of Northwestern Polytechnical University(Grant No.KCJG202432)2023 Northwestern Polytechnical University Degree and Postgraduate Education Research Funding(Grant No.2023YMs014)。
文摘With the rapid development of China’s economy,the demand for high-end talents in the field of engineering technology is becoming increasingly prominent.Engineering doctors,as an important force in this field,have a direct impact on the progress of national technological innovation and the upgrading of industrial structure.Currently,there are still some issues in the university-enterprise collaboration for engineering doctor training in China,such as unclear cooperation mechanisms and responsibility divisions,insufficient corporate participation and enthusiasm,and imperfect evaluation and feedback mechanisms.This paper aims to explore the university-enterprise collaborative training path of engineering doctors from the perspective of deep integration,analyzing multiple dimensions including training objectives,curriculum design,practical sessions,mentor teams,and evaluation systems,in order to provide reference for the reform and practice of engineering doctor training in China.
基金Liaoning Provincial Social Science Planning Fund“Research on the Educational Intelligent Evaluation System Based on the CIPP Model and Artificial Intelligence under the Background of New Engineering”(L22BTJ005)。
文摘This study explores the feasibility of constructing an intelligent educational evaluation system based on the CIPP model and artificial intelligence technology in the context of new engineering disciplines.By integrating the CIPP model with AI technology,a novel intelligent educational evaluation system was designed.Through experimental validation and case studies,the system demonstrated significant effectiveness in improving teaching quality,facilitating personalized student development,and optimizing educational resource allocation.Additionally,the study predicts potential changes this system could bring to the education industry and proposes relevant policy recommendations.Although the current research has limitations,with technological advancements in the future,this system is expected to provide stronger support for innovations in engineering education models.
基金supported by the‘Pioneer’and‘Leading Goose’R&D Program of Zhejiang(Grant No.2023C02018)Zhejiang Provincial Natural Science Foundation of China(Grant No.LTGN23D010002)+2 种基金National Natural Science Foundation of China(Grant No.42371385)Funds of the Natural Science Foundation of Hangzhou(Grant No.2024SZRYBD010001)Nanxun Scholars Program of ZJWEU(Grant No.RC2022010755).
文摘Plant diseases are a major threat that can severely impact the production of agriculture and forestry.This can lead to the disruption of ecosystem functions and health.With its ability to capture continuous narrow-band spectra,hyperspectral technology has become a crucial tool to monitor crop diseases using remote sensing.However,existing continuous wavelet analysis(CWA)methods suffer from feature redundancy issues,while the continuous wavelet projection algorithm(CWPA),an optimization approach for feature selection,has not been fully validated to monitor plant diseases.This study utilized rice bacterial leaf blight(BLB)as an example by evaluating the performance of four wavelet basis functions-Gaussian2,Mexican hat,Meyer,andMorlet-within theCWAandCWPAframeworks.Additionally,the classification models were constructed using the k-nearest neighbors(KNN),randomforest(RF),and Naïve Bayes(NB)algorithms.The results showed the following:(1)Compared to traditional CWA,CWPA significantly reduced the number of required features.Under the CWPA framework,almost all the model combinations achieved maximum classification accuracy with only one feature.In contrast,the CWA framework required three to seven features.(2)Thechoice of wavelet basis functions markedly affected the performance of themodel.Of the four functions tested,the Meyer wavelet demonstrated the best overall performance in both the CWPA and CWA frameworks.(3)Under theCWPAframework,theMeyer-KNNandMeyer-NBcombinations achieved the highest overall accuracy of 93.75%using just one feature.In contrast,under the CWA framework,the CWA-RF combination achieved comparable accuracy(93.75%)but required six features.This study verified the technical advantages of CWPA for monitoring crop diseases,identified an optimal wavelet basis function selection scheme,and provided reliable technical support to precisely monitor BLB in rice(Oryza sativa).Moreover,the proposed methodological framework offers a scalable approach for the early diagnosis and assessment of plant stress,which can contribute to improved accuracy and timeliness when plant stress is monitored.
基金supported by Guizhou Provincial Basic Research Program(Natural Science,Grant No.QKHJC-ZK[2022]YB075)the National Natural Science Foundation of China(Grant No.42067046)+2 种基金the Guizhou Provincial Program on Commercialization of Scientific and Technological Achievements(N0.QKHCG-LH2024-ZD025)the Science and Technology Planning Project of Guiyang City(Grant No.ZKHT[2023]13-10)Undergraduate Training Program for Innovation and Entrepreneurship of Guizhou Province(Project No.S202110657053)。
文摘Rockfall kinematic characteristics exhibit significant randomness and are influenced by factors such as rock mass properties,slope morphology,impact angle,and slope materials.Accurately determining the key parameters of rockfall movement is critical for understanding motion patterns and effectively preventing and controlling rockfall hazards.In this study,a monitoring system consisting of selfdeveloped inertial navigation equipment,high-speed cameras,and an unmanned aerial vehicle was used to conduct onsite motion tests involving four differently shaped rock specimens on three types of slopes(bedrock,detritus,and clast bedding).The selfdeveloped inertial navigation system integrated a highdynamic-range accelerometer(±400 g)and a shockresistant gyroscope(±4000°/s),capable of robustly collecting data during the test.The data collected from these tests were processed to extract key kinematic parameters such as velocity,trajectory,restitution coefficients,and friction coefficients.The test results demonstrated that the inertial navigation system accurately recorded the acceleration and angular velocity of the rocks during motion,with these measurements closely aligning with the field data.The normal and tangential restitution coefficients were found to be influenced primarily by the slope material and impact angle,with higher normal restitution coefficients observed for low-angle impacts.The normal restitution coefficients ranged from 0.35 to 0.86,whereas the tangential restitution coefficients ranged from 0.46 to 0.91,depending on the slope materials.Additionally,the sliding friction coefficient was calculated to be between 0.66 and 0.78,whereas the rolling friction coefficient for the slab-shaped specimen was determined to be 0.53.These findings provide valuable data for improving the accuracy of rockfall trajectory predictions and the design of protective structures.
基金supported by the National Natural Science Foundation of China under Grant No.12204062the Natural Science Foundation of Shandong Province under Grant No.ZR2022MF330。
文摘To enhance speech emotion recognition capability,this study constructs a speech emotion recognition model integrating the adaptive acoustic mixup(AAM)and improved coordinate and shuffle attention(ICASA)methods.The AAM method optimizes data augmentation by combining a sample selection strategy and dynamic interpolation coefficients,thus enabling information fusion of speech data with different emotions at the acoustic level.The ICASA method enhances feature extraction capability through dynamic fusion of the improved coordinate attention(ICA)and shuffle attention(SA)techniques.The ICA technique reduces computational overhead by employing depth-separable convolution and an h-swish activation function and captures long-range dependencies of multi-scale time-frequency features using the attention weights.The SA technique promotes feature interaction through channel shuffling,which helps the model learn richer and more discriminative emotional features.Experimental results demonstrate that,compared to the baseline model,the proposed model improves the weighted accuracy by 5.42%and 4.54%,and the unweighted accuracy by 3.37%and 3.85%on the IEMOCAP and RAVDESS datasets,respectively.These improvements were confirmed to be statistically significant by independent samples t-tests,further supporting the practical reliability and applicability of the proposed model in real-world emotion-aware speech systems.
文摘In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and recognition accuracy on resource-constrained devices such as mobile terminals remains a major challenge.To address this,we propose a novel time-frequency dual-branch parallel residual network,which integrates a Dual-Branch Broadcast Residual module and a Time-Frequency Coordinate Attention module.The time-domain and frequency-domain branches are designed in parallel to independently extract temporal and spectral features,effectively avoiding the potential information loss caused by serial stacking,while enhancing information flow and multi-scale feature fusion.In terms of training strategy,a curriculum learning approach is introduced to progressively improve model robustness fromeasy to difficult tasks.Experimental results demonstrate that the proposed method consistently outperforms existing lightweight models under various signal-to-noise ratio(SNR)conditions,achieving superior far-field recognition performance on the Google Speech Commands V2 dataset.Notably,the model maintains stable performance even in low-SNR environments such as–10 dB,and generalizes well to unseen SNR conditions during training,validating its robustness to novel noise scenarios.Furthermore,the proposed model exhibits significantly fewer parameters,making it highly suitable for deployment on resource-limited devices.Overall,the model achieves a favorable balance between performance and parameter efficiency,demonstrating strong potential for practical applications.
基金TheNationalNatural Science Foundation ofChina(Nos.62272418,62102058)Zhejiang Provincial Natural Science Foundation Major Project(No.LD24F020004)the Major Open Project of Key Laboratory for Advanced Design and Intelligent Computing of the Ministry of Education(No.ADIC2023ZD001).
文摘With the rapid development of smart manufacturing,intelligent safety monitoring in industrial workshops has become increasingly important.To address the challenges of complex backgrounds,target scale variation,and excessive model parameters in worker violation detection,this study proposes ADCP-YOLO,an enhanced lightweight model based on YOLOv8.Here,“ADCP”represents four key improvements:Alterable Kernel Convolution(AKConv),Dilated-Wise Residual(DWR)module,Channel Reconstruction Global Attention Mechanism(CRGAM),and Powerful-IoU loss.These components collaboratively enhance feature extraction,multi-scale perception,and localization accuracy while effectively reducing model complexity and computational cost.Experimental results show that ADCP-YOLO achieves a mAP of 90.6%,surpassing YOLOv8 by 3.0%with a 6.6%reduction in parameters.These findings demonstrate that ADCP-YOLO successfully balances accuracy and efficiency,offering a practical solution for intelligent safety monitoring in smart factory workshops.
文摘The colorectal cancer is one of the most common and lethal cancers,and colorectal polyps,as precancerous lesions,can lead to diagnostic oversight or misdiagnosis due to their varied shapes and sizes,thereby promoting the irreversible progression of colorectal cancer.We propose a YOLO based model and name it EF-YOLO.It incorporates transformer to extract contextual information about the colorectal polyps.Simultaneously,leveraging the morphological characteristics of colorectal polyps,we design a brand-new module,namely advanced multi-scale aggregation(AMSA),to replace the traditional multi-scale module.The backbone adopts deformable convolutional network-maxpool(DCN-MP)to enhance feature extraction while adaptively sampling points to better match the shapes of colorectal polyps.By combining coordinate attention(CA),this model maximizes the use of positional and channel information,more effectively extracting features of colorectal polyps,directing the model’s attention toward the colorectal polyp region.EF-YOLO has made advancement on the merged Kvasir-SEG and CVC-ClinicDB dataset.Compared to the original model,the mean average precision(mAP)of EF-YOLO increases and reaches 96.60%,meeting automated colorectal polyp detection requirements.
文摘In non-independent and identically distributed(non-IID)data environments,model performance often degrades significantly.To address this issue,two improvement methods are proposed:FedReg and FedReg^(*).FedReg is a method based on hybrid regularization aimed at enhancing federated learning in non-IID scenarios.It introduces hybrid regularization to replace traditional L2 regularization,combining the advantages of L1 and L2 regularization to enable feature selection while preventing overfitting.This method better adapts to the diverse data distributions of different clients,improving the overall model performance.FedReg^(*)combines hybrid regularization with weighted model aggregation.In addition to the benefits of hybrid regularization,FedReg^(*)applies a weighted averaging method in the model aggregation process,calculating weights based on the cosine similarity between each client gradient and the global gradient to more reasonably distribute client contributions.By considering variations in data quality and quantity among clients,FedReg^(*)highlights the importance of key clients and enhances the model’s generalization performance.These improvement methods enhance model accuracy and communication efficiency.
基金supported in part by the Guangdong Province Education Science Planning Project(Higher Education Project,Project No.2024GXJK410)Shenzhen Education Science“14th Five-Year Plan”2023 Annual Project on Artificial Intelligence Special Project under Grant No.rgzn23001,the Guangdong Province Higher Education Research and Reform Project under Grant No.YueJiaoGaoHan(2024)No.9+1 种基金the Guangdong Province General Colleges and Universities Innovation Team Project under No.2022KCXTD038the Guangdong Provincial Hardware and System Teaching&Research Office Quality Engineering Project under No.HITSZERP22002.
文摘This study presents a teaching reform for the Object-oriented Software Construction(OOSC)course by integrating outcome-based education(OBE)and the BOPPPS(bridge-In,objectives,pre-assessment,participatory learning,post-assessment,summary)instructional model.The reform addresses the gap between syntax-based programming instruction and the need for higher-level skills in abstraction,modularity,and software architecture.The course is anchored in a semester-long,project-based learning platform centered on a Java-based Aircraft Battle Game,progressing through six iterative experiments.Each experiment targets specific competencies within the structured BOPPPS teaching cycle and is aligned with specific OBE learning outcomes.A case study on the Factory Pattern illustrates how the BOPPPS model fosters conceptual understanding and practical application.Evaluation results from the 2023 and 2024 spring semesters show improved outcomes:Project completion rose from 87%to 95%,37%of students implemented innovative features,and average final grades increased by 7%.The results affirm that the OBE+BOPPPS integration strengthens engagement,deepens understanding,and equips students with real-world software development competencies.
基金supported in part by the by Chongqing Research Program of Basic Research and Frontier Technology under Grant CSTB2025NSCQ-GPX1309.
文摘Small object detection has been a focus of attention since the emergence of deep learning-based object detection.Although classical object detection frameworks have made significant contributions to the development of object detection,there are still many issues to be resolved in detecting small objects due to the inherent complexity and diversity of real-world visual scenes.In particular,the YOLO(You Only Look Once)series of detection models,renowned for their real-time performance,have undergone numerous adaptations aimed at improving the detection of small targets.In this survey,we summarize the state-of-the-art YOLO-based small object detection methods.This review presents a systematic categorization of YOLO-based approaches for small-object detection,organized into four methodological avenues,namely attention-based feature enhancement,detection-head optimization,loss function,and multi-scale feature fusion strategies.We then examine the principal challenges addressed by each category.Finally,we analyze the performance of thesemethods on public benchmarks and,by comparing current approaches,identify limitations and outline directions for future research.
基金supported by the National Natural Science Foundation of China(Grant Nos.62572057,62272049,U24A20331)Beijing Natural Science Foundation(Grant Nos.4232026,4242020)Academic Research Projects of Beijing Union University(Grant No.ZK10202404).
文摘Traffic sign detection is a critical component of driving systems.Single-stage network-based traffic sign detection algorithms,renowned for their fast detection speeds and high accuracy,have become the dominant approach in current practices.However,in complex and dynamic traffic scenes,particularly with smaller traffic sign objects,challenges such as missed and false detections can lead to reduced overall detection accuracy.To address this issue,this paper proposes a detection algorithm that integrates edge and shape information.Recognizing that traffic signs have specific shapes and distinct edge contours,this paper introduces an edge feature extraction branch within the backbone network,enabling adaptive fusion with features of the same hierarchical level.Additionally,a shape prior convolution module is designed to replaces the first two convolutional modules of the backbone network,aimed at enhancing the model's perception ability for specific shape objects and reducing its sensitivity to background noise.The algorithm was evaluated on the CCTSDB and TT100k datasets,and compared to YOLOv8s,the mAP50 values increased by 3.0%and 10.4%,respectively,demonstrating the effectiveness of the proposed method in improving the accuracy of traffic sign detection.
文摘War rehearsals have become increasingly important in national security due to the growing complexity of international affairs.However,traditional rehearsal methods,such as military chess simulations,are inefficient and inflexible,with particularly pronounced limitations in command and decision-making.The overwhelming volume of information and high decision complexity hinder the realization of autonomous and agile command and control.To address this challenge,an intelligent warfare simulation framework named Command-Agent is proposed,which deeply integrates large language models(LLMs)with digital twin battlefields.By constructing a highly realistic battlefield environment through real-time simulation and multi-source data fusion,the natural language interaction capabilities of LLMs are leveraged to lower the command threshold and to enable autonomous command through the Observe-Orient-Decide-Act(OODA)feedback loop.Within the Command-Agent framework,a multimodel collaborative architecture is further adopted to decouple the decision-generation and command-execution functions of LLMs.By combining specialized models such as Deep Seek-R1 and MCTool,the limitations of single-model capabilities are overcome.MCTool is a lightweight execution model fine-tuned for military Function Calling tasks.The framework also introduces a Vector Knowledge Base to mitigate hallucinations commonly exhibited by LLMs.Experimental results demonstrate that Command-Agent not only enables natural language-driven simulation and control but also deeply understands commander intent.Leveraging the multi-model collaborative architecture,during red-blue UAV confrontations involving 2 to 8 UAVs,the integrated score is improved by an average of 41.8%compared to the single-agent system(MCTool),accompanied by a 161.8%optimization in the battle loss ratio.Furthermore,when compared with multi-agent systems lacking the knowledge base,the inclusion of the Vector Knowledge Base further improves overall performance by 16.8%.In comparison with the general model(Qwen2.5-7B),the fine-tuned MCTool leads by 5%in execution efficiency.Therefore,the proposed Command-Agent introduces a novel perspective to the military command system and offers a feasible solution for intelligent battlefield decision-making.
文摘Amid the increasing demand for data sharing,the need for flexible,secure,and auditable access control mechanisms has garnered significant attention in the academic community.However,blockchain-based ciphertextpolicy attribute-based encryption(CP-ABE)schemes still face cumbersome ciphertext re-encryption and insufficient oversight when handling dynamic attribute changes and cross-chain collaboration.To address these issues,we propose a dynamic permission attribute-encryption scheme for multi-chain collaboration.This scheme incorporates a multiauthority architecture for distributed attribute management and integrates an attribute revocation and granting mechanism that eliminates the need for ciphertext re-encryption,effectively reducing both computational and communication overhead.It leverages the InterPlanetary File System(IPFS)for off-chain data storage and constructs a cross-chain regulatory framework—comprising a Hyperledger Fabric business chain and a FISCO BCOS regulatory chain—to record changes in decryption privileges and access behaviors in an auditable manner.Security analysis shows selective indistinguishability under chosen-plaintext attack(sIND-CPA)security under the decisional q-Parallel Bilinear Diffie-Hellman Exponent Assumption(q-PBDHE).In the performance and experimental evaluations,we compared the proposed scheme with several advanced schemes.The results show that,while preserving security,the proposed scheme achieves higher encryption/decryption efficiency and lower storage overhead for ciphertexts and keys.
基金supported by the National Natural Science Foundation of China under Grant 62472264the Natural Science Distinguished Youth Foundation of Shandong Province under Grant ZR2025QA13.
文摘Workflow scheduling is critical for efficient cloud resource management.This paper proposes Tunicate Swarm-Highest Response Ratio Next,a novel scheduler that synergistically combines the Tunicate Swarm Algorithm with the Highest Response Ratio Next policy.The Tunicate Swarm Algorithm generates a cost-minimizing task-to-VM mapping scheme,while the Highest Response Ratio Next dynamically dispatches tasks in the ready queue with the highest-priority.Experimental results demonstrate that the Tunicate Swarm-Highest Response RatioNext reduces costs by up to 94.8%compared to meta-heuristic baselines.It also achieves competitive cost efficiency vs.a learning-based method while offering superior operational simplicity and efficiency,establishing it as a highly practical solution for dynamic cloud environments.
基金supported by the National Key Research and Development Program of China(2020YFB1005704).
文摘Since the advent of smart contracts,security vulnerabilities have remained a persistent challenge,compromsing both the reliability of contract execution and the overall stability of the virtual currency market.Consequently,the academic community has devoted increasing attention to these security risks.However,conventional approaches to vulnerability detection frequently exhibit limited accuracy.To address this limitation,the present study introduces a novel vulnerability detection framework called GNNSE that integrates symbolic execution with graph neural networks(GNNs).The proposedmethod first constructs semantic graphs to comprehensively capture the control flow and data flow dependencies within smart contracts.These graphs are subsequently processed using GNNs to efficiently identify contracts with a high likelihood of vulnerabilities.For these high-risk contracts,symbolic execution is employed to perform fine-grained,path-level analysis,thereby improving overall detection precision.Experimental results on a dataset comprising 10,079 contracts demonstrate that the proposed method achieves detection precisions of 93.58% for reentrancy vulnerabilities and 92.73% for timestamp-dependent vulnerabilities.
基金support from the National Key Research and Development Program of China(Grant No.2021YFF0704102)the Chongqing Education Commission Key Project of Science and Technology Research(Grant No.KJZD-K202400610)the Chongqing Natural Science Foundation General Project(Grant No.CSTB2025NSCQ-GPX1263).
文摘Attribute-Based Encryption(ABE)has emerged as a fundamental access control mechanism in data sharing,enabling data owners to define flexible access policies.A critical aspect of ABE is key revocation,which plays a pivotal role in maintaining security.However,existing key revocation mechanisms face two major challenges:(1)High overhead due to ciphertext and key updates,primarily stemming from the reliance on revocation lists during attribute revocation,which increases computation and communication costs.(2)Limited universality,as many attribute revocation mechanisms are tailored to specific ABE constructions,restricting their broader applicability.To address these challenges,we propose LUAR(Lightweight and Universal Attribute Revocation),a novel revocation mechanism that leverages Intel Software Guard Extensions(SGX)while minimizing its inherent limitations.Given SGX’s constrained memory(≈90 MB in a personal computer)and susceptibility to side-channel attacks,we carefully manage its usage to reduce reliance while mitigating potential collusion risks between cloud service providers and users.To evaluate LUAR’s lightweight and universality,we integrate it with the classic BSW07 scheme,which can be seamlessly replaced with other ABE constructions.Experimental results demonstrate that LUAR enables secure attribute revocation with low computation and communication overhead.The processing time within the SGX environment remains stable at approximately 55 ms,regardless of the complexity of access policies,ensuring no additional storage or computational burden on SGX.Compared to the Hardware-based Revocable Attribute-Based Encryption(HR-ABE)scheme(IEEE S&P 2024),LUAR incurs a slightly higher computational cost within SGX;however,the overall time from initiating a data request to obtaining plaintext is shorter.As access policies grow more complex,LUAR’s advantages become increasingly evident,showcasing its superior efficiency and broader applicability.
基金supported by the National Natural Science Foundation of China(Nos.62025203 and 62372419).
文摘Training large-scale deep neural networks(DNNs)is prone to software and hardware failures,with critical failures often requiring full-machine reboots that substantially prolong training.Existing checkpoint-recovery solutions either cannot tolerate such critical failures or suffer from slow checkpointing and recovery due to constrained input/output bandwidth.In this paper,we propose FastCheck,a checkpoint-recovery framework that accelerates checkpointing and recovery through parallel transmission and tailored compression.First,FastCheck partitions checkpoints into shards and leverages multiple nodes for parallel checkpointing and recovery.Second,it further reduces checkpoint size and overhead with delta compression for weights and index compression for momentum.Third,FastCheck employs lightweight and consistent health status maintenance that accurately tracks node health,preventing checkpoint transmission to failed nodes.We implement FastCheck in PyTorch and evaluate it on multiple DNN models against two baselines.Experimental results show that FastCheck reduces the checkpointing time by up to 78.42%and the recovery time by up to 77.41%,while consistently improving efficiency across different training stages.
