Online examinations have become a dominant assessment mode,increasing concerns over academic integrity.To address the critical challenge of detecting cheating behaviours,this study proposes a hybrid deep learning appr...Online examinations have become a dominant assessment mode,increasing concerns over academic integrity.To address the critical challenge of detecting cheating behaviours,this study proposes a hybrid deep learning approach that combines visual detection and temporal behaviour classification.The methodology utilises object detection models—You Only Look Once(YOLOv12),Faster Region-based Convolutional Neural Network(RCNN),and Single Shot Detector(SSD)MobileNet—integrated with classification models such as Convolutional Neural Networks(CNN),Bidirectional Gated Recurrent Unit(Bi-GRU),and CNN-LSTM(Long Short-Term Memory).Two distinct datasets were used:the Online Exam Proctoring(EOP)dataset from Michigan State University and the School of Computer Science,Duy Tan Unievrsity(SCS-DTU)dataset collected in a controlled classroom setting.A diverse set of cheating behaviours,including book usage,unauthorised interaction,internet access,and mobile phone use,was categorised.Comprehensive experiments evaluated the models based on accuracy,precision,recall,training time,inference speed,and memory usage.We evaluate nine detector-classifier pairings under a unified budget and score them via a calibrated harmonic mean of detection and classification accuracies,enabling deployment-oriented selection under latency and memory constraints.Macro-Precision/Recall/F1 and Receiver Operating Characteristic-Area Under the Curve(ROC-AUC)are reported for the top configurations,revealing consistent advantages of object-centric pipelines for fine-grained cheating cues.The highest overall score is achieved by YOLOv12+CNN(97.15%accuracy),while SSD-MobileNet+CNN provides the best speed-efficiency trade-off for edge devices.This research provides valuable insights into selecting and deploying appropriate deep learning models for maintaining exam integrity under varying resource constraints.展开更多
Background: COVID-19 limitations have hindered the implementation of new technologies by preventing proctors from coming to the site.We share our first experience of magnetic resonance imaging(MRI)-guided focused ultr...Background: COVID-19 limitations have hindered the implementation of new technologies by preventing proctors from coming to the site.We share our first experience of magnetic resonance imaging(MRI)-guided focused ultrasound(MRgFUS)treatment with an international remote online proctorship,and develop and evaluate the methodology of remote MRgFUS proctorship.Methods: This single-center,nonrandomized controlled prospective study included 94 patients:27 with essential tremor(ET)and 67 with tremor-dominant Parkinson's disease(PD).The coming of proctors was impossible,so we arranged for the remote participation of proctors from the United Kingdom,Spain,and Israel.A total of 38 patients(40.4%)received telemedicine-proctored treatment(proctor group)and 56 received their treatment independently(solo group).We used the Clinical Rating Scale for Tremor(CRST)for ET patients and the Unified Parkinson's Disease Rating Scale(UPDRS)Part III for PD patients.Results: In patients with ET,success rates were 81.8%(proctor group)and 100%(solo group)(p=0.22).CRST reduction on the treated side was 71.43%[65.83%;80.56%](proctor group)versus 60.87%[53.99;79.58](solo group)(p=0.19).None of the patients showed worsening of tremors within 1 year.In patients with PD,the success rates were 92.6%(proctor group)and 100%(solo group)(p=0.08).The UPDRS Part III improvement was 30.1%(proctor group)versus 39.9%(solo group)(p=0.003).The 1-year recurrence rate was 40%(proctor group)and 17.5%(solo group)(p=0.04).No complications were observed at 6 months.Conclusions: We developed a feasible and safe methodology for telemedicine remote online-proctored MRgFUS treatment.No significant difference was observed between the solo and developed remote proctor protocols in terms of complication rate,effect,and long-term results;however,UPDRS Part III improvement was better in the PD solo group.This study demonstrated that the MRgFUS international proctorship can be performed successfully remotely.展开更多
The performance of roller compacted concrete(RCC)was greatly influenced by variations in material proportion,optimum moisture content,density of mixes and methodology adopted making it different from conventional conc...The performance of roller compacted concrete(RCC)was greatly influenced by variations in material proportion,optimum moisture content,density of mixes and methodology adopted making it different from conventional concrete mixes.Even though RCC has gained popularity,the complex phenomenon involved in developing the RCC mixes limits it from large-scale applications.In this study,reclaimed asphalt pavement(RAP)incorporated roller-compacted geopolymer concrete(RGC)mixes were developed herein with different compaction techniques such as vibratory hammer(VH),modified proctor(MP),vibration table(VT)and compression machine(CM)are studied and compared with control mixes of natural aggregates.Initially,the effect of alkali solutions such as sodium hydroxide(SH)and sodium silicate(SS)on the physical properties.During,the second phase mechanical properties such as dry density,compressive,flexural and split-tensile strength,modulus of elasticity and microstructure properties will be investigated.The test results revealed that compaction efforts were greatly influenced by the alkali solution.Furthermore,the poor bond characteristics between RAP and the binder matrix had a significant effect on strength properties.Also,the various compaction techniques affected the mechanical properties of mixes developed herein.In Comparison with various compaction efforts,VH and MP produced comparable results,whereas the VT method underestimated and overestimated the various strength properties.Although,the CM method reports comparable results but difficult to maintain consistency in strength aspects.Therefore,optimization of various parameters influencing the concrete properties needs to be achieved for field density.展开更多
The rapid shift to online education has introduced significant challenges to maintaining academic integrity in remote assessments,as traditional proctoring methods fall short in preventing cheating.The increase in che...The rapid shift to online education has introduced significant challenges to maintaining academic integrity in remote assessments,as traditional proctoring methods fall short in preventing cheating.The increase in cheating during online exams highlights the need for efficient,adaptable detection models to uphold academic credibility.This paper presents a comprehensive analysis of various deep learning models for cheating detection in online proctoring systems,evaluating their accuracy,efficiency,and adaptability.We benchmark several advanced architectures,including EfficientNet,MobileNetV2,ResNet variants and more,using two specialized datasets(OEP and OP)tailored for online proctoring contexts.Our findings reveal that EfficientNetB1 and YOLOv5 achieve top performance on the OP dataset,with EfficientNetB1 attaining a peak accuracy of 94.59% and YOLOv5 reaching a mean average precision(mAP@0.5)of 98.3%.For the OEP dataset,ResNet50-CBAM,YOLOv5 and EfficientNetB0 stand out,with ResNet50-CBAMachieving an accuracy of 93.61% and EfficientNetB0 showing robust detection performance with balanced accuracy and computational efficiency.These results underscore the importance of selectingmodels that balance accuracy and efficiency,supporting scalable,effective cheating detection in online assessments.展开更多
It had been suggested to use the West African Compaction Test Procedure since the early 1950’s so as to determine the CBR of gravel lateritic soils in West African countries [1]. This test procedure called West Afric...It had been suggested to use the West African Compaction Test Procedure since the early 1950’s so as to determine the CBR of gravel lateritic soils in West African countries [1]. This test procedure called West African Compaction (WAC) [2] is largely used in road construction in West African countries and had the particularity to be long and use a large amount of material. This note is the result of several comparisons between test procedures taken to determine the CBR from the WAC method to the standardized laboratory test commonly used to determine the CBR.展开更多
文摘Online examinations have become a dominant assessment mode,increasing concerns over academic integrity.To address the critical challenge of detecting cheating behaviours,this study proposes a hybrid deep learning approach that combines visual detection and temporal behaviour classification.The methodology utilises object detection models—You Only Look Once(YOLOv12),Faster Region-based Convolutional Neural Network(RCNN),and Single Shot Detector(SSD)MobileNet—integrated with classification models such as Convolutional Neural Networks(CNN),Bidirectional Gated Recurrent Unit(Bi-GRU),and CNN-LSTM(Long Short-Term Memory).Two distinct datasets were used:the Online Exam Proctoring(EOP)dataset from Michigan State University and the School of Computer Science,Duy Tan Unievrsity(SCS-DTU)dataset collected in a controlled classroom setting.A diverse set of cheating behaviours,including book usage,unauthorised interaction,internet access,and mobile phone use,was categorised.Comprehensive experiments evaluated the models based on accuracy,precision,recall,training time,inference speed,and memory usage.We evaluate nine detector-classifier pairings under a unified budget and score them via a calibrated harmonic mean of detection and classification accuracies,enabling deployment-oriented selection under latency and memory constraints.Macro-Precision/Recall/F1 and Receiver Operating Characteristic-Area Under the Curve(ROC-AUC)are reported for the top configurations,revealing consistent advantages of object-centric pipelines for fine-grained cheating cues.The highest overall score is achieved by YOLOv12+CNN(97.15%accuracy),while SSD-MobileNet+CNN provides the best speed-efficiency trade-off for edge devices.This research provides valuable insights into selecting and deploying appropriate deep learning models for maintaining exam integrity under varying resource constraints.
文摘Background: COVID-19 limitations have hindered the implementation of new technologies by preventing proctors from coming to the site.We share our first experience of magnetic resonance imaging(MRI)-guided focused ultrasound(MRgFUS)treatment with an international remote online proctorship,and develop and evaluate the methodology of remote MRgFUS proctorship.Methods: This single-center,nonrandomized controlled prospective study included 94 patients:27 with essential tremor(ET)and 67 with tremor-dominant Parkinson's disease(PD).The coming of proctors was impossible,so we arranged for the remote participation of proctors from the United Kingdom,Spain,and Israel.A total of 38 patients(40.4%)received telemedicine-proctored treatment(proctor group)and 56 received their treatment independently(solo group).We used the Clinical Rating Scale for Tremor(CRST)for ET patients and the Unified Parkinson's Disease Rating Scale(UPDRS)Part III for PD patients.Results: In patients with ET,success rates were 81.8%(proctor group)and 100%(solo group)(p=0.22).CRST reduction on the treated side was 71.43%[65.83%;80.56%](proctor group)versus 60.87%[53.99;79.58](solo group)(p=0.19).None of the patients showed worsening of tremors within 1 year.In patients with PD,the success rates were 92.6%(proctor group)and 100%(solo group)(p=0.08).The UPDRS Part III improvement was 30.1%(proctor group)versus 39.9%(solo group)(p=0.003).The 1-year recurrence rate was 40%(proctor group)and 17.5%(solo group)(p=0.04).No complications were observed at 6 months.Conclusions: We developed a feasible and safe methodology for telemedicine remote online-proctored MRgFUS treatment.No significant difference was observed between the solo and developed remote proctor protocols in terms of complication rate,effect,and long-term results;however,UPDRS Part III improvement was better in the PD solo group.This study demonstrated that the MRgFUS international proctorship can be performed successfully remotely.
文摘The performance of roller compacted concrete(RCC)was greatly influenced by variations in material proportion,optimum moisture content,density of mixes and methodology adopted making it different from conventional concrete mixes.Even though RCC has gained popularity,the complex phenomenon involved in developing the RCC mixes limits it from large-scale applications.In this study,reclaimed asphalt pavement(RAP)incorporated roller-compacted geopolymer concrete(RGC)mixes were developed herein with different compaction techniques such as vibratory hammer(VH),modified proctor(MP),vibration table(VT)and compression machine(CM)are studied and compared with control mixes of natural aggregates.Initially,the effect of alkali solutions such as sodium hydroxide(SH)and sodium silicate(SS)on the physical properties.During,the second phase mechanical properties such as dry density,compressive,flexural and split-tensile strength,modulus of elasticity and microstructure properties will be investigated.The test results revealed that compaction efforts were greatly influenced by the alkali solution.Furthermore,the poor bond characteristics between RAP and the binder matrix had a significant effect on strength properties.Also,the various compaction techniques affected the mechanical properties of mixes developed herein.In Comparison with various compaction efforts,VH and MP produced comparable results,whereas the VT method underestimated and overestimated the various strength properties.Although,the CM method reports comparable results but difficult to maintain consistency in strength aspects.Therefore,optimization of various parameters influencing the concrete properties needs to be achieved for field density.
基金funded by the Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2025R752),Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘The rapid shift to online education has introduced significant challenges to maintaining academic integrity in remote assessments,as traditional proctoring methods fall short in preventing cheating.The increase in cheating during online exams highlights the need for efficient,adaptable detection models to uphold academic credibility.This paper presents a comprehensive analysis of various deep learning models for cheating detection in online proctoring systems,evaluating their accuracy,efficiency,and adaptability.We benchmark several advanced architectures,including EfficientNet,MobileNetV2,ResNet variants and more,using two specialized datasets(OEP and OP)tailored for online proctoring contexts.Our findings reveal that EfficientNetB1 and YOLOv5 achieve top performance on the OP dataset,with EfficientNetB1 attaining a peak accuracy of 94.59% and YOLOv5 reaching a mean average precision(mAP@0.5)of 98.3%.For the OEP dataset,ResNet50-CBAM,YOLOv5 and EfficientNetB0 stand out,with ResNet50-CBAMachieving an accuracy of 93.61% and EfficientNetB0 showing robust detection performance with balanced accuracy and computational efficiency.These results underscore the importance of selectingmodels that balance accuracy and efficiency,supporting scalable,effective cheating detection in online assessments.
文摘It had been suggested to use the West African Compaction Test Procedure since the early 1950’s so as to determine the CBR of gravel lateritic soils in West African countries [1]. This test procedure called West African Compaction (WAC) [2] is largely used in road construction in West African countries and had the particularity to be long and use a large amount of material. This note is the result of several comparisons between test procedures taken to determine the CBR from the WAC method to the standardized laboratory test commonly used to determine the CBR.
