Medical image segmentation is of critical importance in the domain of contemporary medical imaging.However,U-Net and its variants exhibit limitations in capturing complex nonlinear patterns and global contextual infor...Medical image segmentation is of critical importance in the domain of contemporary medical imaging.However,U-Net and its variants exhibit limitations in capturing complex nonlinear patterns and global contextual information.Although the subsequent U-KAN model enhances nonlinear representation capabilities,it still faces challenges such as gradient vanishing during deep network training and spatial detail loss during feature downsampling,resulting in insufficient segmentation accuracy for edge structures and minute lesions.To address these challenges,this paper proposes the RE-UKAN model,which innovatively improves upon U-KAN.Firstly,a residual network is introduced into the encoder to effectively mitigate gradient vanishing through cross-layer identity mappings,thus enhancing modelling capabilities for complex pathological structures.Secondly,Efficient Local Attention(ELA)is integrated to suppress spatial detail loss during downsampling,thereby improving the perception of edge structures and minute lesions.Experimental results on four public datasets demonstrate that RE-UKAN outperforms existing medical image segmentation methods across multiple evaluation metrics,with particularly outstanding performance on the TN-SCUI 2020 dataset,achieving IoU of 88.18%and Dice of 93.57%.Compared to the baseline model,it achieves improvements of 3.05%and 1.72%,respectively.These results fully demonstrate RE-UKAN’s superior detail retention capability and boundary recognition accuracy in complex medical image segmentation tasks,providing a reliable solution for clinical precision segmentation.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
Recent advances in deep learning have significantly improved image deblurring;however,existing approaches still suffer from limited global context modeling,inadequate detail restoration,and poor texture or edge percep...Recent advances in deep learning have significantly improved image deblurring;however,existing approaches still suffer from limited global context modeling,inadequate detail restoration,and poor texture or edge perception,especially under complex dynamic blur.To address these challenges,we propose the Multi-Resolution Fusion Network(MRFNet),a blind multi-scale deblurring framework that integrates progressive residual connectivity for hierarchical feature fusion.The network employs a three-stage design:(1)TransformerBlocks capture long-range dependencies and reconstruct coarse global structures;(2)Nonlinear Activation Free Blocks(NAFBlocks)enhance local detail representation and mid-level feature fusion;and(3)an optimized residual subnetwork based on gated feature modulation refines texture and edge details for high-fidelity restoration.Extensive experiments demonstrate that MRFNet achieves superior performance compared to state-of-the-art methods.On GoPro,it attains 32.52 dB Peak Signal-to-Noise Ratio(PSNR)and 0.071 Learned Perceptual Image Patch Similarity(LPIPS),outperforming MIMOWNet(32.50 dB,0.075).On HIDE,it achieves 30.25 dB PSNR and 0.945 Structural Similarity Index Measure(SSIM),representing gains of+0.26 dB and+0.015 SSIM over MIMO-UNet(29.99 dB,0.930).On RealBlur-J,it reaches 28.82 dB PSNR and 0.872 SSIM,surpassing MIMO-UNet by+1.19 dB and+0.035 SSIM(27.63 dB,0.837).These results validate the effectiveness of the proposed progressive residual fusion and hybrid attention mechanisms in balancing global context understanding and local detail recovery for blind image deblurring.展开更多
To address the inefficient utilization of electrolytic manganese residue(EMR)caused by its high inert content,this study developed a multifunctional solid waste cementitious material by replacing 50-60%of ordinary Por...To address the inefficient utilization of electrolytic manganese residue(EMR)caused by its high inert content,this study developed a multifunctional solid waste cementitious material by replacing 50-60%of ordinary Portland cement(PO 42.5)with wet-ground electrolytic manganese residue(WEMR),wetground granulated blast-furnace slag(WGBFS),and carbide slag(CS).The mechanical properties,hydration characteristics,microstructure,and carbon emissions of the material were systematically investigated with varying WEMR dosages.The experimental results demonstrates that the wet-grinding process significantly refines the particle size and enhances the reactivity of both EMR and GBFS.As the WEMR dosage increases,the 28-day compressive strength initially rise and then declines.Optimal mechanical performance was achieved with 24%WEMR and 6%CS,yielding a 28-day compressive strength of 48.2 MPa.Advanced analytical techniques,including XRD,TG-DTG,SEM,and MIP,were employed to examine the hydration products.The findings reveal that the wet-grinding-alkali-sulfur synergistic activation system in the multi-solid waste cementitious material effectively utilize EMR to generate abundant hydration products such as AFt and C-(A)-S-H.Additionally,the fine particles of WEMR fill the pores in the mortar,further enhancing compressive strength.The cost and carbon emissions of this multifunctional system are only 65.97%and 46.9% of those of PO 42.5,respectively.This study provides a feasible approach for the efficient utilization of EMR,contributing to sustainable construction practices.展开更多
High-resolution remote sensing imagery is essential for critical applications such as precision agriculture,urban management planning,and military reconnaissance.Although significant progress has been made in singleim...High-resolution remote sensing imagery is essential for critical applications such as precision agriculture,urban management planning,and military reconnaissance.Although significant progress has been made in singleimage super-resolution(SISR)using generative adversarial networks(GANs),existing approaches still face challenges in recovering high-frequency details,effectively utilizing features,maintaining structural integrity,and ensuring training stability—particularly when dealing with the complex textures characteristic of remote sensing imagery.To address these limitations,this paper proposes the Improved ResidualModule and AttentionMechanism Network(IRMANet),a novel architecture specifically designed for remote sensing image reconstruction.IRMANet builds upon the Super-Resolution Generative Adversarial Network(SRGAN)framework and introduces several key innovations.First,the Enhanced Residual Unit(ERU)enhances feature reuse and stabilizes training through deep residual connections.Second,the Self-Attention Residual Block(SARB)incorporates a self-attentionmechanism into the Improved Residual Module(IRM)to effectivelymodel long-range dependencies and automatically emphasize salient features.Additionally,the IRM adopts amulti-scale feature fusion strategy to facilitate synergistic interactions between local detail and global semantic information.The effectiveness of each component is validated through ablation studies,while comprehensive comparative experiments on standard remote sensing datasets demonstrate that IRMANet significantly outperforms both the baseline and state-of-the-art methods in terms of perceptual quality and quantitative metrics.Specifically,compared to the baseline model,at a magnification factor of 2,IRMANet achieves an improvement of 0.24 dB in peak signal-to-noise ratio(PSNR)and 0.54 in structural similarity index(SSIM);at a magnification factor of 4,it achieves gains of 0.22 dB in PSNR and 0.51 in SSIM.These results confirm that the proposedmethod effectively enhances detail representation and structural reconstruction accuracy in complex remote sensing scenarios,offering robust technical support for high-precision detection and identification of both military and civilian aircraft.展开更多
Aiming at the problem of potential information noise introduced during the generation of ghost feature maps in GhostNet,this paper proposes a novel lightweight neural network model called ResghostNet.This model constr...Aiming at the problem of potential information noise introduced during the generation of ghost feature maps in GhostNet,this paper proposes a novel lightweight neural network model called ResghostNet.This model constructs the Resghost Module by combining residual connections and Adaptive-SE Blocks,which enhances the quality of generated feature maps through direct propagation of original input information and selection of important channels before cheap operations.Specifically,ResghostNet introduces residual connections on the basis of the Ghost Module to optimize the information flow,and designs a weight self-attention mechanism combined with SE blocks to enhance feature expression capabilities in cheap operations.Experimental results on the ImageNet dataset show that,compared to GhostNet,ResghostNet achieves higher accuracy while reducing the number of parameters by 52%.Although the computational complexity increases,by optimizing the usage strategy of GPU cachememory,themodel’s inference speed becomes faster.The ResghostNet is optimized in terms of classification accuracy and the number of model parameters,and shows great potential in edge computing devices.展开更多
Rapidly improving infertile croplands and enhancing their soil organic carbon(SOC)pool necessitate substantial organic materials incorporation.Converting loose crop straw into granulated form facilitates uniform incor...Rapidly improving infertile croplands and enhancing their soil organic carbon(SOC)pool necessitate substantial organic materials incorporation.Converting loose crop straw into granulated form facilitates uniform incorporation within the plough soil layer.As an innovative soil amelioration approach,the efficiency and patterns of SOC accumulation remain unclear.Two field experiments were conducted in infertile subtropical upland and paddy soils with 0,30,60,and 90 Mg ha^(-1)granulated straw incorporation.After one year,SOC accumulation efficiency from straw input remained stable in upland(30.8–37.5%)with increasing amounts of straw incorporation,while declined from 60.0 to 38.3%in paddy.In both croplands,the contributions of lignin phenols to SOC increased with increasing straw incorporation,while the contributions from amino sugars remained constant at higher straw input levels.Subsequently,the ratios of lignin phenols to amino sugars increased with increasing straw incorporation,indicating faster plant residue accumulation compared to microbial necromass,as the granulation approach limited microbial involvement in straw transformation.Thus,single-time incorporation of substantial granulated straw presents an effective agricultural strategy for rapid amelioration of infertile croplands.展开更多
The curing behavior of composites significantly influences their performance,making it crucial to understand the curing process.This study experimentally measured specific heat capacity,thermal conductivity,glass tran...The curing behavior of composites significantly influences their performance,making it crucial to understand the curing process.This study experimentally measured specific heat capacity,thermal conductivity,glass transition temperature,coefficient of thermal expansion,and cure shrinkage of materials.A simulation model of its curing deformation was established and validated against strain data obtained from fiber Bragg grating experiments.The effects of thickness,heating rate,and cooling rate on the curing temperature field and residual stress field during the molding of thick-section composite plates were analyzed.展开更多
Statistical distribution of residual fatigue life(RFL)of railway axles under given loading was computed using the Monte Carlo method by considering random variation of the selected input parameters.Experimental data f...Statistical distribution of residual fatigue life(RFL)of railway axles under given loading was computed using the Monte Carlo method by considering random variation of the selected input parameters.Experimental data for the EA4T railway axle steel,the loading spectrum,the press fit loading and the residual stress induced by surface hardening were considered in the crack propagation simulations.Usually,the material properties measured by tensile tests are considered to be the most informative source of material data.Under fatigue loading,however,the crack growth rates near the threshold are the most critical data.Two important influencing factors on these crack growth rates are presented:first,the air humidity and,second,the near-surface residual stress.The typical variation of these parameters in operation may change the RFL by one or two orders of magnitude.Experimentally obtained crack growth thresholds and residual stress profiles are highly affected by the used methodology.Therefore,the obtained input data may be located anywhere within a large scatter,while the experimenters are completely unaware of it.This can lead to dangerously non-conservative situations,e.g.when the thresholds are measured in a laboratory under humid air conditions and then applied to predictions of RFLs of axles operated in winter in low air humidity.This is significant for the topic of inspection interval optimisation.The results of experiments done on real 1:1 railway axles were close to the most frequent value found in the histogram of the numerically computed RFLs.展开更多
The neutron diffusion equation plays a pivotal role in nuclear reactor analysis.Nevertheless,employing the physics-informed neural network(PINN)method for its solution entails certain limitations.Conventional PINN app...The neutron diffusion equation plays a pivotal role in nuclear reactor analysis.Nevertheless,employing the physics-informed neural network(PINN)method for its solution entails certain limitations.Conventional PINN approaches generally utilize a fully connected network(FCN)architecture that is susceptible to overfitting,training instability,and gradient vanishing as the network depth increases.These challenges result in accuracy bottlenecks in the solution.In response to these issues,the residual-based resample physics-informed neural network(R2-PINN)is proposed.It is an improved PINN architecture that replaces the FCN with a convolutional neural network with a shortcut(S-CNN).It incorporates skip connections to facilitate gradient propagation between network layers.Additionally,the incorporation of the residual adaptive resampling(RAR)mechanism dynamically increases the number of sampling points.This,in turn,enhances the spatial representation capabilities and overall predictive accuracy of the model.The experimental results illustrate that our approach significantly improves the convergence capability of the model and achieves high-precision predictions of the physical fields.Compared with conventional FCN-based PINN methods,R 2-PINN effectively overcomes the limitations inherent in current methods.Thus,it provides more accurate and robust solutions for neutron diffusion equations.展开更多
This letter introduces the novel concept of Painlevé solitons—waves arising from the interaction between Painlevé waves and solitons in integrable systems.Painlevé solitons can also be viewed as solito...This letter introduces the novel concept of Painlevé solitons—waves arising from the interaction between Painlevé waves and solitons in integrable systems.Painlevé solitons can also be viewed as solitons propagating against a Painlevé wave background,in analogy to the established notion of elliptic solitons,which refers to solitons on an elliptic wave background.By employing a novel symmetry decomposition method aided by nonlocal residual symmetries,we explicitly construct (extended) Painlevé Ⅱ solitons for the Korteweg-de Vries equation and (extended) Painlevé Ⅳ solitons for the Boussinesq equation.展开更多
This study proposes a multi-scale simplified residual convolutional neural network(MS-SRCNN)for the precise prediction of Mg-Nd binary alloy compositions from scanning electron microscope(SEM)images.A multi-scale data...This study proposes a multi-scale simplified residual convolutional neural network(MS-SRCNN)for the precise prediction of Mg-Nd binary alloy compositions from scanning electron microscope(SEM)images.A multi-scale data structure is established by spatially aligning and stacking SEM images at different magnifications.The MS-SRCNN significantly reduces computational runtime by over 90%compared to traditional architectures like ResNet50,VGG16,and VGG19,without compromising prediction accuracy.The model demonstrates more excellent predictive performance,achieving a>5%increase in R^(2) compared to single-scale models.Furthermore,the MS-SRCNN exhibits robust composition prediction capability across other Mg-based binary alloys,including Mg-La,Mg-Sn,Mg-Ce,Mg-Sm,Mg-Ag,and Mg-Y,thereby emphasizing its generalization and extrapolation potential.This research establishes a non-destructive,microstructure-informed composition analysis framework,reduces characterization time compared to traditional experiment methods and provides insights into the composition-microstructure relationship in diverse material systems.展开更多
With growing concerns regarding electromagnetic pollution,low-cost,environmentally friendly,and high-performance electromagnetic wave absorption(EWA)materials have attracted significant attention.This paper reports on...With growing concerns regarding electromagnetic pollution,low-cost,environmentally friendly,and high-performance electromagnetic wave absorption(EWA)materials have attracted significant attention.This paper reports on the synthesis of porous Fe_(3)O_(4)/C composites that incorporate dielectric and magnetic loss mechanisms via the carbothermal reduction method and optimization of waste ratio to enhance EWA performance.The Fe_(3)O_(4)/C composites with 10wt%soybean residues(Fe_(3)O_(4)/C-10),demonstrated the best EWA performance,achieving the minimum reflection loss of−56.4 dB and a bandwidth of 2.14 GHz at a thickness of 2.23 mm.This enhanced EWA performance is primarily attributable to improved impedance matching and the synergistic effect between dielectric and magnetic losses.Furthermore,radar cross-sectional simulations confirmed the practical feasibility of the porous Fe_(3)O_(4)/C composites.This study proposes a viable strategy for utilizing soybean residue and electrolytic manganese residue,highlighting their potential applications in EWA.展开更多
Despite the superior advantages of specific emitter identification in extracting emitter features from in-phase and quadrature(I/Q)signals,challenges persist due to signal-type confusion and background noise interfere...Despite the superior advantages of specific emitter identification in extracting emitter features from in-phase and quadrature(I/Q)signals,challenges persist due to signal-type confusion and background noise interference.To address those limitations,this paper proposes a multi-channel contrast prediction coding and complex-valued residuals network(MCPC-MCVResNet)framework.This model employs contrast prediction techniques to directly extract discriminative features from electromagnetic signal sequences,effectively capturing both amplitude and phase information within I/Q data.A core innovation of this approach is the sphere space softmax(SS-softmax)loss,which optimizes intra-class clustering density of while establishing well-defined boundaries between distinct emitters.The SS-softmax mechanism significantly enhances the model's capacity to discern subtle variations among radiation emitters.Experimental results demonstrate superior identification accuracy,rapid convergence,and exceptional robustness in low signal-to-noise ratio environments.展开更多
The biodegradable polybutylene succinate(PBS)material offers a sustainable solution for a circular economy to address the global issue of marine plastic waste.Its cross-linkage with non-biodegradable xanthan gum(XG)bi...The biodegradable polybutylene succinate(PBS)material offers a sustainable solution for a circular economy to address the global issue of marine plastic waste.Its cross-linkage with non-biodegradable xanthan gum(XG)biopolymer to ameliorate residual granitic soil(RGS)in arid and semiarid regions can significantly mitigate soil erosion.This study investigates the enhancement of RGS by cross-linking the PBS and XG biopolymers.Employing a multitude of geotechnical tests(liquid limit,linear shrinkage,specific gravity,compaction,and UCS tests)at 3 d,28 d,and 90 d of steam-curing at a controlled temperature of 16℃,the outcomes were validated through scanning electron microscopy(SEM),thermogravimetric analysis(TGA),Fourier transform infrared spectroscopy(FTIR),and Brunauer-Emmett-Teller(BET)analyses.In addition,a comprehensive experimental database of 150 tests and nine parameters from the current study was utilized to model the UCS90-d(i.e.unconfined compressive strength after 90 d of curing)of the PBS-XG-treated RGS mixtures by deploying the random forest(RF)and eXtreme Gradient Boost(XGBoost)methods.The results found that the two biopolymers significantly improve the mechanical properties of RGS,with optimal UCS achieved at specific dosages(0.4PBS,1.5XG,and 0.2PBS+1.5XG dosage levels)and curing times.The UCS of PBS-XG-treated RGS showed up to a 57%increase after 90 d of curing.Furthermore,SEM and FTIR analyses revealed the formation of stronger microstructures and chemical bonds,respectively,whereas BET analysis indicated that pore volume and diameter are critical in affecting UCS.The proposed RF model outperformed XGBoost in predictive accuracy and generalization,demonstrating robustness and versatility.Moreover,SHAP values highlighted the significant impact of input parameters on UCS90-d,with curing time and specific material properties being key determinants.The study concludes with the proposal of a novel PyCharm intuitive graphical user interface as a"UCS Prediction App"for engineers and practitioners to forecast the UCS90-d of granitic residual soil.展开更多
A new manufactured soil product (Turba) was produced using acidified bauxite residue into which 10% green waste compost had been incorporated. A laboratory/greenhouse experiment was carried out to determine if sand co...A new manufactured soil product (Turba) was produced using acidified bauxite residue into which 10% green waste compost had been incorporated. A laboratory/greenhouse experiment was carried out to determine if sand could be used as an ingredient or an amendment for Turba. Sand was added at rates of 0%, 5%, 10%, 25, 50% and 75% (w/w) in two different ways 1) by incorporating it into the Turba during its manufacture (IN) or 2) by mixing it with Turba aggregates after their manufacture (OUT). Incorporation of sand into Turba aggregates (IN) decreased the percentage of sample present as large aggregates (2 - 4 mm dia.) after crushing and sieving (<4 mm) and also reduced the stability of 2 - 4 mm dia. formed aggregates (to dry/wet sieving) and are therefore not recommended. In a 16-week greenhouse study, ryegrass shoot yields were greater in Turba than in sand [and decreased with increasing sand additions (OUT)] while root dry matter showed the opposite trend. The greater grass growth in Turba than sand was attributed to incipit water stress in plants grown in sand and this may have promoted greater allocation of assimilates to roots resulting in a greater root-to-top mass ratio. The much lower macroporosity in Turba coupled with the solid cemented nature of Turba aggregates resulted in production of thinner roots and therefore greater root length than in sand. Turba (manufactured from bauxite residue and compost added at 10% w/w) is a suitable medium for plant growth and there is no advantage in incorporating sand into, or with, the Turba aggregates.展开更多
Erythromycin fermentation residue(EFR)represents a typical hazardous waste produced by the microbial pharmaceutical industry.Although electrolysis is promising for EFR disposal,its microbial threats remain unclear.Her...Erythromycin fermentation residue(EFR)represents a typical hazardous waste produced by the microbial pharmaceutical industry.Although electrolysis is promising for EFR disposal,its microbial threats remain unclear.Herein,metagenomics was coupled with the random forest technique to decipher the antibiotic resistance patterns of electrochemically treated EFR.Results showed that 95.75%of erythromycin could be removed in 2 hr.Electrolysis temporarily influenced EFRmicrobiota,where the relative abundances of Proteobacteria and Actinobacteria increased,while those of Fusobacteria,Firmicutes,and Bacteroidetes decreased.A total of 505 antibiotic resistance gene(ARG)subtypes encoding resistance to 21 antibiotic types and 150 mobile genetic elements(MGEs),mainly including plasmid(72)and transposase(52)were assembled in EFR.Significant linear regression models were identified among microbial richness,ARG subtypes,and MGE numbers(r^(2)=0.50-0.81,p<0.001).Physicochemical factors of EFR(Total nitrogen,total organic carbon,protein,and humus)regulated ARG and MGE assembly(%IncMSE value=5.14-14.85).The core ARG,MGE,and microbe sets(93.08%-99.85%)successfully explained 89.71%-92.92%of total ARG and MGE abundances.Specifically,gene aph(3 )-I,transposase tnpA,and Mycolicibacterium were the primary drivers of the resistance dissemination system.This study also proposes efficient resistance mitigation measures,and provides recommendations for future management of antibiotic fermentation residue.展开更多
In the aerospace field, residual stress directly affects the strength, fatigue life and dimensional stability of thin-walled structural components, and is a key factor to ensure flight safety and reliability. At prese...In the aerospace field, residual stress directly affects the strength, fatigue life and dimensional stability of thin-walled structural components, and is a key factor to ensure flight safety and reliability. At present, research on residual stress at home and abroad mainly focuses on the optimization of traditional detection technology, stress control of manufacturing process and service performance evaluation, among which research on residual stress detection methods mainly focuses on the improvement of the accuracy, sensitivity, reliability and other performance of existing detection methods, but it still faces many challenges such as extremely small detection range, low efficiency, large error and limited application range.展开更多
Residual neural network (ResNet) is a powerful neural network architecture that has proven to be excellent in extracting spatial and channel-wise information of images. ResNet employs a residual learning strategy that...Residual neural network (ResNet) is a powerful neural network architecture that has proven to be excellent in extracting spatial and channel-wise information of images. ResNet employs a residual learning strategy that maps inputs directly to outputs, making it less difficult to optimize. In this paper, we incorporate differential information into the original residual block to improve the representative ability of the ResNet, allowing the modified network to capture more complex and metaphysical features. The proposed DFNet preserves the features after each convolutional operation in the residual block, and combines the feature maps of different levels of abstraction through the differential information. To verify the effectiveness of DFNet on image recognition, we select six distinct classification datasets. The experimental results show that our proposed DFNet has better performance and generalization ability than other state-of-the-art variants of ResNet in terms of classification accuracy and other statistical analysis.展开更多
文摘Medical image segmentation is of critical importance in the domain of contemporary medical imaging.However,U-Net and its variants exhibit limitations in capturing complex nonlinear patterns and global contextual information.Although the subsequent U-KAN model enhances nonlinear representation capabilities,it still faces challenges such as gradient vanishing during deep network training and spatial detail loss during feature downsampling,resulting in insufficient segmentation accuracy for edge structures and minute lesions.To address these challenges,this paper proposes the RE-UKAN model,which innovatively improves upon U-KAN.Firstly,a residual network is introduced into the encoder to effectively mitigate gradient vanishing through cross-layer identity mappings,thus enhancing modelling capabilities for complex pathological structures.Secondly,Efficient Local Attention(ELA)is integrated to suppress spatial detail loss during downsampling,thereby improving the perception of edge structures and minute lesions.Experimental results on four public datasets demonstrate that RE-UKAN outperforms existing medical image segmentation methods across multiple evaluation metrics,with particularly outstanding performance on the TN-SCUI 2020 dataset,achieving IoU of 88.18%and Dice of 93.57%.Compared to the baseline model,it achieves improvements of 3.05%and 1.72%,respectively.These results fully demonstrate RE-UKAN’s superior detail retention capability and boundary recognition accuracy in complex medical image segmentation tasks,providing a reliable solution for clinical precision segmentation.
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funded by Qinghai University Postgraduate Research and Practice Innovation Program of Funder,grant number 2025-GMKY-42.
文摘Recent advances in deep learning have significantly improved image deblurring;however,existing approaches still suffer from limited global context modeling,inadequate detail restoration,and poor texture or edge perception,especially under complex dynamic blur.To address these challenges,we propose the Multi-Resolution Fusion Network(MRFNet),a blind multi-scale deblurring framework that integrates progressive residual connectivity for hierarchical feature fusion.The network employs a three-stage design:(1)TransformerBlocks capture long-range dependencies and reconstruct coarse global structures;(2)Nonlinear Activation Free Blocks(NAFBlocks)enhance local detail representation and mid-level feature fusion;and(3)an optimized residual subnetwork based on gated feature modulation refines texture and edge details for high-fidelity restoration.Extensive experiments demonstrate that MRFNet achieves superior performance compared to state-of-the-art methods.On GoPro,it attains 32.52 dB Peak Signal-to-Noise Ratio(PSNR)and 0.071 Learned Perceptual Image Patch Similarity(LPIPS),outperforming MIMOWNet(32.50 dB,0.075).On HIDE,it achieves 30.25 dB PSNR and 0.945 Structural Similarity Index Measure(SSIM),representing gains of+0.26 dB and+0.015 SSIM over MIMO-UNet(29.99 dB,0.930).On RealBlur-J,it reaches 28.82 dB PSNR and 0.872 SSIM,surpassing MIMO-UNet by+1.19 dB and+0.035 SSIM(27.63 dB,0.837).These results validate the effectiveness of the proposed progressive residual fusion and hybrid attention mechanisms in balancing global context understanding and local detail recovery for blind image deblurring.
基金Funded by the Guangxi Key Research and Development Program(Nos.GK AB24010020,and GK AB23026071)the Key Project of Guangxi Natural Science Foundation(No.2025GXNSFDA090046)the Guangxi Science and Technology Base and Talent Special Project(No.GK AD24010062)。
文摘To address the inefficient utilization of electrolytic manganese residue(EMR)caused by its high inert content,this study developed a multifunctional solid waste cementitious material by replacing 50-60%of ordinary Portland cement(PO 42.5)with wet-ground electrolytic manganese residue(WEMR),wetground granulated blast-furnace slag(WGBFS),and carbide slag(CS).The mechanical properties,hydration characteristics,microstructure,and carbon emissions of the material were systematically investigated with varying WEMR dosages.The experimental results demonstrates that the wet-grinding process significantly refines the particle size and enhances the reactivity of both EMR and GBFS.As the WEMR dosage increases,the 28-day compressive strength initially rise and then declines.Optimal mechanical performance was achieved with 24%WEMR and 6%CS,yielding a 28-day compressive strength of 48.2 MPa.Advanced analytical techniques,including XRD,TG-DTG,SEM,and MIP,were employed to examine the hydration products.The findings reveal that the wet-grinding-alkali-sulfur synergistic activation system in the multi-solid waste cementitious material effectively utilize EMR to generate abundant hydration products such as AFt and C-(A)-S-H.Additionally,the fine particles of WEMR fill the pores in the mortar,further enhancing compressive strength.The cost and carbon emissions of this multifunctional system are only 65.97%and 46.9% of those of PO 42.5,respectively.This study provides a feasible approach for the efficient utilization of EMR,contributing to sustainable construction practices.
基金funded by the Henan Province Key R&D Program Project,“Research and Application Demonstration of Class Ⅱ Superlattice Medium Wave High Temperature Infrared Detector Technology”,grant number 231111210400.
文摘High-resolution remote sensing imagery is essential for critical applications such as precision agriculture,urban management planning,and military reconnaissance.Although significant progress has been made in singleimage super-resolution(SISR)using generative adversarial networks(GANs),existing approaches still face challenges in recovering high-frequency details,effectively utilizing features,maintaining structural integrity,and ensuring training stability—particularly when dealing with the complex textures characteristic of remote sensing imagery.To address these limitations,this paper proposes the Improved ResidualModule and AttentionMechanism Network(IRMANet),a novel architecture specifically designed for remote sensing image reconstruction.IRMANet builds upon the Super-Resolution Generative Adversarial Network(SRGAN)framework and introduces several key innovations.First,the Enhanced Residual Unit(ERU)enhances feature reuse and stabilizes training through deep residual connections.Second,the Self-Attention Residual Block(SARB)incorporates a self-attentionmechanism into the Improved Residual Module(IRM)to effectivelymodel long-range dependencies and automatically emphasize salient features.Additionally,the IRM adopts amulti-scale feature fusion strategy to facilitate synergistic interactions between local detail and global semantic information.The effectiveness of each component is validated through ablation studies,while comprehensive comparative experiments on standard remote sensing datasets demonstrate that IRMANet significantly outperforms both the baseline and state-of-the-art methods in terms of perceptual quality and quantitative metrics.Specifically,compared to the baseline model,at a magnification factor of 2,IRMANet achieves an improvement of 0.24 dB in peak signal-to-noise ratio(PSNR)and 0.54 in structural similarity index(SSIM);at a magnification factor of 4,it achieves gains of 0.22 dB in PSNR and 0.51 in SSIM.These results confirm that the proposedmethod effectively enhances detail representation and structural reconstruction accuracy in complex remote sensing scenarios,offering robust technical support for high-precision detection and identification of both military and civilian aircraft.
基金funded by Science and Technology Innovation Project grant No.ZZKY20222304.
文摘Aiming at the problem of potential information noise introduced during the generation of ghost feature maps in GhostNet,this paper proposes a novel lightweight neural network model called ResghostNet.This model constructs the Resghost Module by combining residual connections and Adaptive-SE Blocks,which enhances the quality of generated feature maps through direct propagation of original input information and selection of important channels before cheap operations.Specifically,ResghostNet introduces residual connections on the basis of the Ghost Module to optimize the information flow,and designs a weight self-attention mechanism combined with SE blocks to enhance feature expression capabilities in cheap operations.Experimental results on the ImageNet dataset show that,compared to GhostNet,ResghostNet achieves higher accuracy while reducing the number of parameters by 52%.Although the computational complexity increases,by optimizing the usage strategy of GPU cachememory,themodel’s inference speed becomes faster.The ResghostNet is optimized in terms of classification accuracy and the number of model parameters,and shows great potential in edge computing devices.
基金financially supported by the National Key R&D Program of China(2021YFD1901203 and 2021YFD1901204)the Strategic Priority Research Program of the Chinese Academy of Sciences(XDA0440404)+2 种基金the National Natural Science Foundation of China(42377348)the Science Foundation for Distinguished Young Scholars of Hunan Province,China(2024JJ2052)the Natural Science Foundation of Guangxi,China(2025GXNSFAA069337)。
文摘Rapidly improving infertile croplands and enhancing their soil organic carbon(SOC)pool necessitate substantial organic materials incorporation.Converting loose crop straw into granulated form facilitates uniform incorporation within the plough soil layer.As an innovative soil amelioration approach,the efficiency and patterns of SOC accumulation remain unclear.Two field experiments were conducted in infertile subtropical upland and paddy soils with 0,30,60,and 90 Mg ha^(-1)granulated straw incorporation.After one year,SOC accumulation efficiency from straw input remained stable in upland(30.8–37.5%)with increasing amounts of straw incorporation,while declined from 60.0 to 38.3%in paddy.In both croplands,the contributions of lignin phenols to SOC increased with increasing straw incorporation,while the contributions from amino sugars remained constant at higher straw input levels.Subsequently,the ratios of lignin phenols to amino sugars increased with increasing straw incorporation,indicating faster plant residue accumulation compared to microbial necromass,as the granulation approach limited microbial involvement in straw transformation.Thus,single-time incorporation of substantial granulated straw presents an effective agricultural strategy for rapid amelioration of infertile croplands.
基金supported by the National Natural Science Foundation of China(Grant Nos.12172045,U2241240,and 12221002)the National Program on Key Basic Research Project,China(Grant No.2019-JCJQ-ZD-308-00).
文摘The curing behavior of composites significantly influences their performance,making it crucial to understand the curing process.This study experimentally measured specific heat capacity,thermal conductivity,glass transition temperature,coefficient of thermal expansion,and cure shrinkage of materials.A simulation model of its curing deformation was established and validated against strain data obtained from fiber Bragg grating experiments.The effects of thickness,heating rate,and cooling rate on the curing temperature field and residual stress field during the molding of thick-section composite plates were analyzed.
基金financially supported by the Czech Science Foundation in the frame of the project No.22-28283Sby the Technology Agency of the Czech Republic through the project No.CK03000060.
文摘Statistical distribution of residual fatigue life(RFL)of railway axles under given loading was computed using the Monte Carlo method by considering random variation of the selected input parameters.Experimental data for the EA4T railway axle steel,the loading spectrum,the press fit loading and the residual stress induced by surface hardening were considered in the crack propagation simulations.Usually,the material properties measured by tensile tests are considered to be the most informative source of material data.Under fatigue loading,however,the crack growth rates near the threshold are the most critical data.Two important influencing factors on these crack growth rates are presented:first,the air humidity and,second,the near-surface residual stress.The typical variation of these parameters in operation may change the RFL by one or two orders of magnitude.Experimentally obtained crack growth thresholds and residual stress profiles are highly affected by the used methodology.Therefore,the obtained input data may be located anywhere within a large scatter,while the experimenters are completely unaware of it.This can lead to dangerously non-conservative situations,e.g.when the thresholds are measured in a laboratory under humid air conditions and then applied to predictions of RFLs of axles operated in winter in low air humidity.This is significant for the topic of inspection interval optimisation.The results of experiments done on real 1:1 railway axles were close to the most frequent value found in the histogram of the numerically computed RFLs.
基金supported by the Science and Technology on Reactor System Design Technology Laboratory(No.LRSDT12023108)supported in part by the Chongqing Postdoctoral Science Foundation(No.cstc2021jcyj-bsh0252)+2 种基金the National Natural Science Foundation of China(No.12005030)Sichuan Province to unveil the list of marshal industry common technology research projects(No.23jBGOV0001)Special Program for Stabilizing Support to Basic Research of National Basic Research Institutes(No.WDZC-2023-05-03-05).
文摘The neutron diffusion equation plays a pivotal role in nuclear reactor analysis.Nevertheless,employing the physics-informed neural network(PINN)method for its solution entails certain limitations.Conventional PINN approaches generally utilize a fully connected network(FCN)architecture that is susceptible to overfitting,training instability,and gradient vanishing as the network depth increases.These challenges result in accuracy bottlenecks in the solution.In response to these issues,the residual-based resample physics-informed neural network(R2-PINN)is proposed.It is an improved PINN architecture that replaces the FCN with a convolutional neural network with a shortcut(S-CNN).It incorporates skip connections to facilitate gradient propagation between network layers.Additionally,the incorporation of the residual adaptive resampling(RAR)mechanism dynamically increases the number of sampling points.This,in turn,enhances the spatial representation capabilities and overall predictive accuracy of the model.The experimental results illustrate that our approach significantly improves the convergence capability of the model and achieves high-precision predictions of the physical fields.Compared with conventional FCN-based PINN methods,R 2-PINN effectively overcomes the limitations inherent in current methods.Thus,it provides more accurate and robust solutions for neutron diffusion equations.
基金supported by the National Natural Science Foundations of China (Grant Nos.12235007,12001424,12271324,and 12501333)the Natural Science Basic research program of Shaanxi Province (Grant Nos.2021JZ-21 and 2024JC-YBQN-0069)+3 种基金the China Postdoctoral Science Foundation (Grant Nos.2020M673332 and 2024M751921)the Fundamental Research Funds for the Central Universities (Grant No.GK202304028)the 2023 Shaanxi Province Postdoctoral Research Project (Grant No.2023BSHEDZZ186)Xi’an University,Xi’an Science and Technology Plan Wutongshu Technology Transfer Action Innovation Team(Grant No.25WTZD07)。
文摘This letter introduces the novel concept of Painlevé solitons—waves arising from the interaction between Painlevé waves and solitons in integrable systems.Painlevé solitons can also be viewed as solitons propagating against a Painlevé wave background,in analogy to the established notion of elliptic solitons,which refers to solitons on an elliptic wave background.By employing a novel symmetry decomposition method aided by nonlocal residual symmetries,we explicitly construct (extended) Painlevé Ⅱ solitons for the Korteweg-de Vries equation and (extended) Painlevé Ⅳ solitons for the Boussinesq equation.
基金funded by the National Natural Science Foundation of China(No.52204407)the Natural Science Foundation of Jiangsu Province(No.BK20220595)the China Postdoctoral Science Foundation(No.2022M723689).
文摘This study proposes a multi-scale simplified residual convolutional neural network(MS-SRCNN)for the precise prediction of Mg-Nd binary alloy compositions from scanning electron microscope(SEM)images.A multi-scale data structure is established by spatially aligning and stacking SEM images at different magnifications.The MS-SRCNN significantly reduces computational runtime by over 90%compared to traditional architectures like ResNet50,VGG16,and VGG19,without compromising prediction accuracy.The model demonstrates more excellent predictive performance,achieving a>5%increase in R^(2) compared to single-scale models.Furthermore,the MS-SRCNN exhibits robust composition prediction capability across other Mg-based binary alloys,including Mg-La,Mg-Sn,Mg-Ce,Mg-Sm,Mg-Ag,and Mg-Y,thereby emphasizing its generalization and extrapolation potential.This research establishes a non-destructive,microstructure-informed composition analysis framework,reduces characterization time compared to traditional experiment methods and provides insights into the composition-microstructure relationship in diverse material systems.
基金supported by the National Natural Science Foundation of China(No.52471221)the Natural Science Foundation of Hunan Province,China(No.2024JJ7145)the National Sustainable Development Agenda Innovation Demonstration Zone Hunan special project,China(No.2022sfq09).
文摘With growing concerns regarding electromagnetic pollution,low-cost,environmentally friendly,and high-performance electromagnetic wave absorption(EWA)materials have attracted significant attention.This paper reports on the synthesis of porous Fe_(3)O_(4)/C composites that incorporate dielectric and magnetic loss mechanisms via the carbothermal reduction method and optimization of waste ratio to enhance EWA performance.The Fe_(3)O_(4)/C composites with 10wt%soybean residues(Fe_(3)O_(4)/C-10),demonstrated the best EWA performance,achieving the minimum reflection loss of−56.4 dB and a bandwidth of 2.14 GHz at a thickness of 2.23 mm.This enhanced EWA performance is primarily attributable to improved impedance matching and the synergistic effect between dielectric and magnetic losses.Furthermore,radar cross-sectional simulations confirmed the practical feasibility of the porous Fe_(3)O_(4)/C composites.This study proposes a viable strategy for utilizing soybean residue and electrolytic manganese residue,highlighting their potential applications in EWA.
基金supported by the National Natural Science Foundation of China(62201602)。
文摘Despite the superior advantages of specific emitter identification in extracting emitter features from in-phase and quadrature(I/Q)signals,challenges persist due to signal-type confusion and background noise interference.To address those limitations,this paper proposes a multi-channel contrast prediction coding and complex-valued residuals network(MCPC-MCVResNet)framework.This model employs contrast prediction techniques to directly extract discriminative features from electromagnetic signal sequences,effectively capturing both amplitude and phase information within I/Q data.A core innovation of this approach is the sphere space softmax(SS-softmax)loss,which optimizes intra-class clustering density of while establishing well-defined boundaries between distinct emitters.The SS-softmax mechanism significantly enhances the model's capacity to discern subtle variations among radiation emitters.Experimental results demonstrate superior identification accuracy,rapid convergence,and exceptional robustness in low signal-to-noise ratio environments.
基金supported by the National Natural Science Foundation of China(Grant Nos.52379104 and 52090084).
文摘The biodegradable polybutylene succinate(PBS)material offers a sustainable solution for a circular economy to address the global issue of marine plastic waste.Its cross-linkage with non-biodegradable xanthan gum(XG)biopolymer to ameliorate residual granitic soil(RGS)in arid and semiarid regions can significantly mitigate soil erosion.This study investigates the enhancement of RGS by cross-linking the PBS and XG biopolymers.Employing a multitude of geotechnical tests(liquid limit,linear shrinkage,specific gravity,compaction,and UCS tests)at 3 d,28 d,and 90 d of steam-curing at a controlled temperature of 16℃,the outcomes were validated through scanning electron microscopy(SEM),thermogravimetric analysis(TGA),Fourier transform infrared spectroscopy(FTIR),and Brunauer-Emmett-Teller(BET)analyses.In addition,a comprehensive experimental database of 150 tests and nine parameters from the current study was utilized to model the UCS90-d(i.e.unconfined compressive strength after 90 d of curing)of the PBS-XG-treated RGS mixtures by deploying the random forest(RF)and eXtreme Gradient Boost(XGBoost)methods.The results found that the two biopolymers significantly improve the mechanical properties of RGS,with optimal UCS achieved at specific dosages(0.4PBS,1.5XG,and 0.2PBS+1.5XG dosage levels)and curing times.The UCS of PBS-XG-treated RGS showed up to a 57%increase after 90 d of curing.Furthermore,SEM and FTIR analyses revealed the formation of stronger microstructures and chemical bonds,respectively,whereas BET analysis indicated that pore volume and diameter are critical in affecting UCS.The proposed RF model outperformed XGBoost in predictive accuracy and generalization,demonstrating robustness and versatility.Moreover,SHAP values highlighted the significant impact of input parameters on UCS90-d,with curing time and specific material properties being key determinants.The study concludes with the proposal of a novel PyCharm intuitive graphical user interface as a"UCS Prediction App"for engineers and practitioners to forecast the UCS90-d of granitic residual soil.
文摘A new manufactured soil product (Turba) was produced using acidified bauxite residue into which 10% green waste compost had been incorporated. A laboratory/greenhouse experiment was carried out to determine if sand could be used as an ingredient or an amendment for Turba. Sand was added at rates of 0%, 5%, 10%, 25, 50% and 75% (w/w) in two different ways 1) by incorporating it into the Turba during its manufacture (IN) or 2) by mixing it with Turba aggregates after their manufacture (OUT). Incorporation of sand into Turba aggregates (IN) decreased the percentage of sample present as large aggregates (2 - 4 mm dia.) after crushing and sieving (<4 mm) and also reduced the stability of 2 - 4 mm dia. formed aggregates (to dry/wet sieving) and are therefore not recommended. In a 16-week greenhouse study, ryegrass shoot yields were greater in Turba than in sand [and decreased with increasing sand additions (OUT)] while root dry matter showed the opposite trend. The greater grass growth in Turba than sand was attributed to incipit water stress in plants grown in sand and this may have promoted greater allocation of assimilates to roots resulting in a greater root-to-top mass ratio. The much lower macroporosity in Turba coupled with the solid cemented nature of Turba aggregates resulted in production of thinner roots and therefore greater root length than in sand. Turba (manufactured from bauxite residue and compost added at 10% w/w) is a suitable medium for plant growth and there is no advantage in incorporating sand into, or with, the Turba aggregates.
基金supported by the Special Project of Basic Scientific Research Business of Central Public Welfare Scientific Research Institutes (No.2019YSKY-027).
文摘Erythromycin fermentation residue(EFR)represents a typical hazardous waste produced by the microbial pharmaceutical industry.Although electrolysis is promising for EFR disposal,its microbial threats remain unclear.Herein,metagenomics was coupled with the random forest technique to decipher the antibiotic resistance patterns of electrochemically treated EFR.Results showed that 95.75%of erythromycin could be removed in 2 hr.Electrolysis temporarily influenced EFRmicrobiota,where the relative abundances of Proteobacteria and Actinobacteria increased,while those of Fusobacteria,Firmicutes,and Bacteroidetes decreased.A total of 505 antibiotic resistance gene(ARG)subtypes encoding resistance to 21 antibiotic types and 150 mobile genetic elements(MGEs),mainly including plasmid(72)and transposase(52)were assembled in EFR.Significant linear regression models were identified among microbial richness,ARG subtypes,and MGE numbers(r^(2)=0.50-0.81,p<0.001).Physicochemical factors of EFR(Total nitrogen,total organic carbon,protein,and humus)regulated ARG and MGE assembly(%IncMSE value=5.14-14.85).The core ARG,MGE,and microbe sets(93.08%-99.85%)successfully explained 89.71%-92.92%of total ARG and MGE abundances.Specifically,gene aph(3 )-I,transposase tnpA,and Mycolicibacterium were the primary drivers of the resistance dissemination system.This study also proposes efficient resistance mitigation measures,and provides recommendations for future management of antibiotic fermentation residue.
文摘In the aerospace field, residual stress directly affects the strength, fatigue life and dimensional stability of thin-walled structural components, and is a key factor to ensure flight safety and reliability. At present, research on residual stress at home and abroad mainly focuses on the optimization of traditional detection technology, stress control of manufacturing process and service performance evaluation, among which research on residual stress detection methods mainly focuses on the improvement of the accuracy, sensitivity, reliability and other performance of existing detection methods, but it still faces many challenges such as extremely small detection range, low efficiency, large error and limited application range.
基金supported by the Japan Society for the Promotion of Science(JSPS)KAKENHI under Grant JP22H03643Japan Science and Technology Agency(JST)Support for Pioneering Research Initiated by the Next Generation(SPRING)under Grant JPMJSP2145JST through the Establishment of University Fellowships towards the Creation of Science Technology Innovation under Grant JPMJFS2115.
文摘Residual neural network (ResNet) is a powerful neural network architecture that has proven to be excellent in extracting spatial and channel-wise information of images. ResNet employs a residual learning strategy that maps inputs directly to outputs, making it less difficult to optimize. In this paper, we incorporate differential information into the original residual block to improve the representative ability of the ResNet, allowing the modified network to capture more complex and metaphysical features. The proposed DFNet preserves the features after each convolutional operation in the residual block, and combines the feature maps of different levels of abstraction through the differential information. To verify the effectiveness of DFNet on image recognition, we select six distinct classification datasets. The experimental results show that our proposed DFNet has better performance and generalization ability than other state-of-the-art variants of ResNet in terms of classification accuracy and other statistical analysis.
