Brain hemorrhage is a serious and life-threatening condition. It cancause permanent and lifelong disability even when it is not fatal. The wordhemorrhage denotes leakage of blood within the brain and this leakage ofbl...Brain hemorrhage is a serious and life-threatening condition. It cancause permanent and lifelong disability even when it is not fatal. The wordhemorrhage denotes leakage of blood within the brain and this leakage ofblood from capillaries causes stroke and adequate supply of oxygen to thebrain is hindered. Modern imaging methods such as computed tomography(CT) and magnetic resonance imaging (MRI) are employed to get an idearegarding the extent of the damage. An early diagnosis and treatment can savelives and limit the adverse effects of a brain hemorrhage. In this case, a deepneural network (DNN) is an effective choice for the early identification andclassification of brain hemorrhage for the timely recovery and treatment of anaffected person. In this paper, the proposed research work is divided into twonovel approaches, where, one for the classification and the other for volumecalculation of brain hemorrhage. Two different datasets are used for twodifferent techniques classification and volume. A novel algorithm is proposedto calculate the volume of hemorrhage using CT scan images. In the firstapproach, the ‘RSNA’ dataset is used to classify the brain hemorrhage typesusing transfer learning and achieved an accuracy of 93.77%. Furthermore,in the second approach, a novel algorithm has been proposed to calculate thevolume of brain hemorrhage and achieved tremendous results as 1035.91mm3and 9.25 cm3, using the PhysioNet CT scan tomography dataset.展开更多
In the light of superior interaction between pyridine unit and perovskite,a facile star-shaped triphenylamine-based hole transport material(HTM)incorporating pyridine core(coded as H-Pyr)is designed and synthesized.A ...In the light of superior interaction between pyridine unit and perovskite,a facile star-shaped triphenylamine-based hole transport material(HTM)incorporating pyridine core(coded as H-Pyr)is designed and synthesized.A reference HTM with benzene core,coded as H-Ben,is also prepared for a comparative study.The effects of varying core on HTMs are investigated by comparing the photophysical,electrochemical and hole mobility properties.It is found that pyridine core exhibits better conjunction and decreased dihedral angles with triphenylamine side arms than that of benzene,leading to obviously better hole mobility and well-matched work function.The perovskite film prepared on H-Pyr also shows improved crystallization than on H-Ben.Photoluminescence and electrochemical impedance studies indicate improved charge extraction and reduced recombination in the H-Pyr-based perovskite solar cells.Consequently,H-Pyr-based device exhibits higher efficiency than H-Ben-based one.After doping with a Lewis acid,tris(pentafluorophenyl)borane,H-Pyr-based device delivers a champion efficiency of 17.09%,which is much higher compared with 12.14% of the device employing conventional poly(3,4-ethy lenedioxythiophene)polystyrene sulfonate(PEDOT:PSS)as HTM.Moreover,the H-Pyr-based device displays good long-term stability that the power conversion efficiency remains over 80% of the initial value after storage in ambient(relative humidity=50±5%)for 20 days.展开更多
Several recent successes in deep learning(DL),such as state-of-the-art performance on several image classification benchmarks,have been achieved through the improved configuration.Hyperparameters(HPs)tuning is a key f...Several recent successes in deep learning(DL),such as state-of-the-art performance on several image classification benchmarks,have been achieved through the improved configuration.Hyperparameters(HPs)tuning is a key factor affecting the performance of machine learning(ML)algorithms.Various state-of-the-art DL models use different HPs in different ways for classification tasks on different datasets.This manuscript provides a brief overview of learning parameters and configuration techniques to show the benefits of using a large-scale handdrawn sketch dataset for classification problems.We analyzed the impact of different learning parameters and toplayer configurations with batch normalization(BN)and dropouts on the performance of the pre-trained visual geometry group 19(VGG-19).The analyzed learning parameters include different learning rates and momentum values of two different optimizers,such as stochastic gradient descent(SGD)and Adam.Our analysis demonstrates that using the SGD optimizer and learning parameters,such as small learning rates with high values of momentum,along with both BN and dropouts in top layers,has a good impact on the sketch image classification accuracy.展开更多
The quality of the perovskite light absorption layer plays a dynamic role in the photovoltaic properties of solar cells.The existing methods to prepare methylammonium lead iodide(MAPbI3)films render substantial struct...The quality of the perovskite light absorption layer plays a dynamic role in the photovoltaic properties of solar cells.The existing methods to prepare methylammonium lead iodide(MAPbI3)films render substantial structural defect density,particularly at the grain boundaries and film surface,constituting a challenge that hinders the further optoelectronic enhancement of perovskite solar cells.Herein,a unique approach was introduced:using a simple ethylammonium chloride(EACl)additive in perovskite precursor mixture to produce high-quality MAPbI3 thin films.The results indicated that EACl could encourage perovskite crystal growth without experiencing the intermediate phase formation and would evaporate from the perovskite after annealing.Additionally,a gradient perovskite structure was achieved using this technique,which impressively enhanced the performance of the perovskite films.A high power conversion efficiency(PCE)of 20.03%was achieved under the optimal amount of EACl,and the resultant efficient device could retain over 89%of the original PCE after aging for 1000 h at room temperature.This novel technique leads to a facile fabrication of highquality and less-defect perovskite thin films for competent and stable devices.展开更多
Bimetallic nanocrystals (BMNCs) with distinguished electronic and chemical properties from those of their parent metals,offer the opportunity to obtain new catalysts with enhanced selectivity,activity,and stability.He...Bimetallic nanocrystals (BMNCs) with distinguished electronic and chemical properties from those of their parent metals,offer the opportunity to obtain new catalysts with enhanced selectivity,activity,and stability.Here we describe the facile synthesis of rhodium-copper bimetallic system with different compositions and uniform morphology for chemo selective hydrogenation of functionalized quinolines.Our findings demonstrate that Rh-Cu BMNCs exhibited composition dependent activity and selectivity.BMNCs with rhodium to copper ratio 3:1 surpassed individual Rh and Cu and other compositions both in activity and selectivity for quinolines hydrogenation and performed even better than Rh/C with same amount of Rh.Rh3Cui catalyst displayed excellent tolerance for synthetically significant functional groups such as -OH,NH2,F,particularly for aldehyde group which is very reactive towards reduction.These results suggested that the coexistence of rhodium and copper metals play important role in the enhancement of catalytic activity due to synergistic effects and revealed that bimetallic nanocrystals can be promising as practical catalysts for selective hydrogenation of quinoline and other substrates.展开更多
Selective hydrogenation of the carbonyl bond inα,β-unsaturated carbonyl compounds is rather challenging owing to the more feasible hydrogenation of ethylenic bond from both thermodynamic and kinetic aspects.Here,we ...Selective hydrogenation of the carbonyl bond inα,β-unsaturated carbonyl compounds is rather challenging owing to the more feasible hydrogenation of ethylenic bond from both thermodynamic and kinetic aspects.Here,we demonstrate a facile emulsionbased molecule-nanoparticle self-assembly strategy for the atomic engineering of Ir species on three-dimensional CeO_(2)spheres(Ir1@CeO_(2)).When applied to the hydrogenation ofα,β-unsaturated aldehydes,Ir1@CeO_(2)catalyst remarkably exhibited~100%selectivity towards unsaturated alcohols,whereas the formation of Ir nanoparticles on CeO_(2)drastically decreased the selectivity for unsaturated alcohols.Spectroscopic studies revealed that strong metal-support interactions triggered the charge transfer from Ir to CeO_(2),leading to the partial reduction of Ce^(4+)to Ce^(3+)along with the formation new Ir^(δ+)-O_(2)--Ce^(3+)(OV)interfaces.The electrophilic atomic Ir species at the Ir^(δ+)-O_(2)--Ce^(3+)(OV)interfaces would therefore preferentially adsorb and facilitate hydrogenation of polar C=O bond to achieve exceptional selectivity.展开更多
As the boom of mobile devices,Android mobile apps play an irreplaceable roles in people’s daily life,which have the characteristics of frequent updates involving in many code commits to meet new requirements.Just-in-...As the boom of mobile devices,Android mobile apps play an irreplaceable roles in people’s daily life,which have the characteristics of frequent updates involving in many code commits to meet new requirements.Just-in-Time(JIT)defect prediction aims to identify whether the commit instances will bring defects into the new release of apps and provides immediate feedback to developers,which is more suitable to mobile apps.As the within-app defect prediction needs sufficient historical data to label the commit instances,which is inadequate in practice,one alternative method is to use the cross-project model.In this work,we propose a novel method,called KAL,for cross-project JIT defect prediction task in the context of Android mobile apps.More specifically,KAL first transforms the commit instances into a high-dimensional feature space using kernel-based principal component analysis technique to obtain the representative features.Then,the adversarial learning technique is used to extract the common feature embedding for the model building.We conduct experiments on 14 Android mobile apps and employ four effort-aware indicators for performance evaluation.The results on 182 cross-project pairs demonstrate that our proposed KAL method obtains better performance than 20 comparative methods.展开更多
Recent studies have shown remarkable success in face image generation task.However,existing approaches have limited diversity,quality and controllability in generating results.To address these issues,we propose a nove...Recent studies have shown remarkable success in face image generation task.However,existing approaches have limited diversity,quality and controllability in generating results.To address these issues,we propose a novel end-to-end learning framework to generate diverse,realistic and controllable face images guided by face masks.The face mask provides a good geometric constraint for a face by specifying the size and location of different components of the face,such as eyes,nose and mouse.The framework consists of four components:style encoder,style decoder,generator and discriminator.The style encoder generates a style code which represents the style of the result face;the generator translate the input face mask into a real face based on the style code;the style decoder learns to reconstruct the style code from the generated face image;and the discriminator classifies an input face image as real or fake.With the style code,the proposed model can generate different face images matching the input face mask,and by manipulating the face mask,we can finely control the generated face image.We empirically demonstrate the effectiveness of our approach on mask guided face image synthesis task.展开更多
文摘Brain hemorrhage is a serious and life-threatening condition. It cancause permanent and lifelong disability even when it is not fatal. The wordhemorrhage denotes leakage of blood within the brain and this leakage ofblood from capillaries causes stroke and adequate supply of oxygen to thebrain is hindered. Modern imaging methods such as computed tomography(CT) and magnetic resonance imaging (MRI) are employed to get an idearegarding the extent of the damage. An early diagnosis and treatment can savelives and limit the adverse effects of a brain hemorrhage. In this case, a deepneural network (DNN) is an effective choice for the early identification andclassification of brain hemorrhage for the timely recovery and treatment of anaffected person. In this paper, the proposed research work is divided into twonovel approaches, where, one for the classification and the other for volumecalculation of brain hemorrhage. Two different datasets are used for twodifferent techniques classification and volume. A novel algorithm is proposedto calculate the volume of hemorrhage using CT scan images. In the firstapproach, the ‘RSNA’ dataset is used to classify the brain hemorrhage typesusing transfer learning and achieved an accuracy of 93.77%. Furthermore,in the second approach, a novel algorithm has been proposed to calculate thevolume of brain hemorrhage and achieved tremendous results as 1035.91mm3and 9.25 cm3, using the PhysioNet CT scan tomography dataset.
基金supported by the National Key R&D Program of China(2019YFB1503202)the 111 Project(B16016)+1 种基金the National Natural Science Foundation of China(61904053,51702096,U1705256 and 51572080)the Fundamental Research Funds for the Central Universities(2019MS026,2019MS027 and 2020MS080)。
文摘In the light of superior interaction between pyridine unit and perovskite,a facile star-shaped triphenylamine-based hole transport material(HTM)incorporating pyridine core(coded as H-Pyr)is designed and synthesized.A reference HTM with benzene core,coded as H-Ben,is also prepared for a comparative study.The effects of varying core on HTMs are investigated by comparing the photophysical,electrochemical and hole mobility properties.It is found that pyridine core exhibits better conjunction and decreased dihedral angles with triphenylamine side arms than that of benzene,leading to obviously better hole mobility and well-matched work function.The perovskite film prepared on H-Pyr also shows improved crystallization than on H-Ben.Photoluminescence and electrochemical impedance studies indicate improved charge extraction and reduced recombination in the H-Pyr-based perovskite solar cells.Consequently,H-Pyr-based device exhibits higher efficiency than H-Ben-based one.After doping with a Lewis acid,tris(pentafluorophenyl)borane,H-Pyr-based device delivers a champion efficiency of 17.09%,which is much higher compared with 12.14% of the device employing conventional poly(3,4-ethy lenedioxythiophene)polystyrene sulfonate(PEDOT:PSS)as HTM.Moreover,the H-Pyr-based device displays good long-term stability that the power conversion efficiency remains over 80% of the initial value after storage in ambient(relative humidity=50±5%)for 20 days.
文摘Several recent successes in deep learning(DL),such as state-of-the-art performance on several image classification benchmarks,have been achieved through the improved configuration.Hyperparameters(HPs)tuning is a key factor affecting the performance of machine learning(ML)algorithms.Various state-of-the-art DL models use different HPs in different ways for classification tasks on different datasets.This manuscript provides a brief overview of learning parameters and configuration techniques to show the benefits of using a large-scale handdrawn sketch dataset for classification problems.We analyzed the impact of different learning parameters and toplayer configurations with batch normalization(BN)and dropouts on the performance of the pre-trained visual geometry group 19(VGG-19).The analyzed learning parameters include different learning rates and momentum values of two different optimizers,such as stochastic gradient descent(SGD)and Adam.Our analysis demonstrates that using the SGD optimizer and learning parameters,such as small learning rates with high values of momentum,along with both BN and dropouts in top layers,has a good impact on the sketch image classification accuracy.
基金supported by the National Key R&D Program of China(2019YFB1503202)the 111 Project(B16016)+1 种基金the National Natural Science Foundation of China(51702096,U1705256 and 61904053)the Fundamental Research Funds for the Central Universities(2019MS026,2019MS027 and 2020MS080)。
文摘The quality of the perovskite light absorption layer plays a dynamic role in the photovoltaic properties of solar cells.The existing methods to prepare methylammonium lead iodide(MAPbI3)films render substantial structural defect density,particularly at the grain boundaries and film surface,constituting a challenge that hinders the further optoelectronic enhancement of perovskite solar cells.Herein,a unique approach was introduced:using a simple ethylammonium chloride(EACl)additive in perovskite precursor mixture to produce high-quality MAPbI3 thin films.The results indicated that EACl could encourage perovskite crystal growth without experiencing the intermediate phase formation and would evaporate from the perovskite after annealing.Additionally,a gradient perovskite structure was achieved using this technique,which impressively enhanced the performance of the perovskite films.A high power conversion efficiency(PCE)of 20.03%was achieved under the optimal amount of EACl,and the resultant efficient device could retain over 89%of the original PCE after aging for 1000 h at room temperature.This novel technique leads to a facile fabrication of highquality and less-defect perovskite thin films for competent and stable devices.
基金supported by the National Key R&D Program of China (Nos.2016YFA0202801 and 2017YFA0700101)the National Natural Science Foundation of China (Nos.21573119,21590792,21890383,and 21872076)China Postdoctoral Science Foundation (No.2018M631444).
文摘Bimetallic nanocrystals (BMNCs) with distinguished electronic and chemical properties from those of their parent metals,offer the opportunity to obtain new catalysts with enhanced selectivity,activity,and stability.Here we describe the facile synthesis of rhodium-copper bimetallic system with different compositions and uniform morphology for chemo selective hydrogenation of functionalized quinolines.Our findings demonstrate that Rh-Cu BMNCs exhibited composition dependent activity and selectivity.BMNCs with rhodium to copper ratio 3:1 surpassed individual Rh and Cu and other compositions both in activity and selectivity for quinolines hydrogenation and performed even better than Rh/C with same amount of Rh.Rh3Cui catalyst displayed excellent tolerance for synthetically significant functional groups such as -OH,NH2,F,particularly for aldehyde group which is very reactive towards reduction.These results suggested that the coexistence of rhodium and copper metals play important role in the enhancement of catalytic activity due to synergistic effects and revealed that bimetallic nanocrystals can be promising as practical catalysts for selective hydrogenation of quinoline and other substrates.
基金the National Natural Science Foundation of China(No.21901007)the Natural Science Foundation of Anhui Province(No.2008085QB83)+1 种基金the Science and Technology Development Fund(FDCT)of Macao SAR(No.0032/2021/ITP)the University of Macao(No.MYRG2020-00026-FST)。
文摘Selective hydrogenation of the carbonyl bond inα,β-unsaturated carbonyl compounds is rather challenging owing to the more feasible hydrogenation of ethylenic bond from both thermodynamic and kinetic aspects.Here,we demonstrate a facile emulsionbased molecule-nanoparticle self-assembly strategy for the atomic engineering of Ir species on three-dimensional CeO_(2)spheres(Ir1@CeO_(2)).When applied to the hydrogenation ofα,β-unsaturated aldehydes,Ir1@CeO_(2)catalyst remarkably exhibited~100%selectivity towards unsaturated alcohols,whereas the formation of Ir nanoparticles on CeO_(2)drastically decreased the selectivity for unsaturated alcohols.Spectroscopic studies revealed that strong metal-support interactions triggered the charge transfer from Ir to CeO_(2),leading to the partial reduction of Ce^(4+)to Ce^(3+)along with the formation new Ir^(δ+)-O_(2)--Ce^(3+)(OV)interfaces.The electrophilic atomic Ir species at the Ir^(δ+)-O_(2)--Ce^(3+)(OV)interfaces would therefore preferentially adsorb and facilitate hydrogenation of polar C=O bond to achieve exceptional selectivity.
基金supported by the National Natural Science Foundation of China (Grant No.62072060).
文摘As the boom of mobile devices,Android mobile apps play an irreplaceable roles in people’s daily life,which have the characteristics of frequent updates involving in many code commits to meet new requirements.Just-in-Time(JIT)defect prediction aims to identify whether the commit instances will bring defects into the new release of apps and provides immediate feedback to developers,which is more suitable to mobile apps.As the within-app defect prediction needs sufficient historical data to label the commit instances,which is inadequate in practice,one alternative method is to use the cross-project model.In this work,we propose a novel method,called KAL,for cross-project JIT defect prediction task in the context of Android mobile apps.More specifically,KAL first transforms the commit instances into a high-dimensional feature space using kernel-based principal component analysis technique to obtain the representative features.Then,the adversarial learning technique is used to extract the common feature embedding for the model building.We conduct experiments on 14 Android mobile apps and employ four effort-aware indicators for performance evaluation.The results on 182 cross-project pairs demonstrate that our proposed KAL method obtains better performance than 20 comparative methods.
基金This work is supported by the National Key Research and Development Program of China(2018YFF0214700).
文摘Recent studies have shown remarkable success in face image generation task.However,existing approaches have limited diversity,quality and controllability in generating results.To address these issues,we propose a novel end-to-end learning framework to generate diverse,realistic and controllable face images guided by face masks.The face mask provides a good geometric constraint for a face by specifying the size and location of different components of the face,such as eyes,nose and mouse.The framework consists of four components:style encoder,style decoder,generator and discriminator.The style encoder generates a style code which represents the style of the result face;the generator translate the input face mask into a real face based on the style code;the style decoder learns to reconstruct the style code from the generated face image;and the discriminator classifies an input face image as real or fake.With the style code,the proposed model can generate different face images matching the input face mask,and by manipulating the face mask,we can finely control the generated face image.We empirically demonstrate the effectiveness of our approach on mask guided face image synthesis task.
