The emergence of next generation networks(NextG),including 5G and beyond,is reshaping the technological landscape of cellular and mobile networks.These networks are sufficiently scaled to interconnect billions of user...The emergence of next generation networks(NextG),including 5G and beyond,is reshaping the technological landscape of cellular and mobile networks.These networks are sufficiently scaled to interconnect billions of users and devices.Researchers in academia and industry are focusing on technological advancements to achieve highspeed transmission,cell planning,and latency reduction to facilitate emerging applications such as virtual reality,the metaverse,smart cities,smart health,and autonomous vehicles.NextG continuously improves its network functionality to support these applications.Multiple input multiple output(MIMO)technology offers spectral efficiency,dependability,and overall performance in conjunctionwithNextG.This article proposes a secure channel estimation technique in MIMO topology using a norm-estimation model to provide comprehensive insights into protecting NextG network components against adversarial attacks.The technique aims to create long-lasting and secure NextG networks using this extended approach.The viability of MIMO applications and modern AI-driven methodologies to combat cybersecurity threats are explored in this research.Moreover,the proposed model demonstrates high performance in terms of reliability and accuracy,with a 20%reduction in the MalOut-RealOut-Diff metric compared to existing state-of-the-art techniques.展开更多
To investigate the overall performance of reverse energy bypass scramjet,firstly a variable spe⁃cific heat method combined with a chemical balance calculation module for combustion products were used to es⁃tablish a b...To investigate the overall performance of reverse energy bypass scramjet,firstly a variable spe⁃cific heat method combined with a chemical balance calculation module for combustion products were used to es⁃tablish a benchmark scramjet performance evaluation model.Based on the test data of typical flying point of Mach 7 with the altitude of 29 km,the reliability of the model was verified.The deviations of parameters such as the to⁃tal pressure loss of combustor between the model and the test data were analyzed.Furtherly,an analytical method for post-combustion magnetohydrodynamic power generation was established;by embedding the above method into the overall performance evaluation model,performance prediction considering the power generation effect was realized.Finally,based on the above model,variety regulations of the inlet and the outlet parameters of the power generation channel and performance parameters including the engine specific impulse and the unit thrust under different enthalpy extraction ratios and load factors were analyzed.It could be concluded that the model can reliably predict the variations of key parameters.As the value of the load factor increases,the value of the conduc⁃tivity required to reach the specified enthalpy extraction ratio first decreases and then increases,which is approxi⁃mately parabolic.In order to reduce the demand for the gas conductivity for MHD power generation,the load fac⁃tor should be around 0.5.When the load factor is 0.4 and the magnetic induction intensity is 2.5 T,if the enthalpy extraction ratio reaches 0.5%,the engine specific impulse performance reduces about 3.58%.展开更多
The generation of optical vortices from nonlinear photonic crystals(NPCs)with spatially modulated second-order nonlinearity offers a promising approach to extend the working wavelength and topological charge of vortex...The generation of optical vortices from nonlinear photonic crystals(NPCs)with spatially modulated second-order nonlinearity offers a promising approach to extend the working wavelength and topological charge of vortex beams for various applications.In this work,the second harmonic(SH)optical vortex beams generated from nonlinear fork gratings under Gaussian beam illumination are numerically investigated.The far-field intensity and phase distributions,as well as the orbital angular momentum(OAM)spectra of the SH beams,are analyzed for different structural topological charges and diffraction orders.Results reveal that higher-order diffraction and larger structural topological charges lead to angular interference patterns and non-uniform intensity distributions,deviating from the standard vortex profile.To optimize the SH vortex quality,the effects of the fundamental wave beam waist,crystal thickness,and grating duty cycle are explored.It is shown that increasing the beam waist can effectively suppress diffraction order interference and improve the beam’s quality.This study provides theoretical guidance for enhancing the performance of nonlinear optical devices based on NPCs.展开更多
The application of generative artificial intelligence(AI)is bringing about notable changes in anime creation.This paper surveys recent advancements and applications of diffusion and language models in anime generation...The application of generative artificial intelligence(AI)is bringing about notable changes in anime creation.This paper surveys recent advancements and applications of diffusion and language models in anime generation,focusing on their demonstrated potential to enhance production efficiency through automation and personalization.Despite these benefits,it is crucial to acknowledge the substantial initial computational investments required for training and deploying these models.We conduct an in-depth survey of cutting-edge generative AI technologies,encompassing models such as Stable Diffusion and GPT,and appraise pivotal large-scale datasets alongside quantifiable evaluation metrics.Review of the surveyed literature indicates the achievement of considerable maturity in the capacity of AI models to synthesize high-quality,aesthetically compelling anime visual images from textual prompts,alongside discernible progress in the generation of coherent narratives.However,achieving perfect long-form consistency,mitigating artifacts like flickering in video sequences,and enabling fine-grained artistic control remain critical ongoing challenges.Building upon these advancements,research efforts have increasingly pivoted towards the synthesis of higher-dimensional content,such as video and three-dimensional assets,with recent studies demonstrating significant progress in this burgeoning field.Nevertheless,formidable challenges endure amidst these advancements.Foremost among these are the substantial computational exigencies requisite for training and deploying these sophisticated models,particularly pronounced in the realm of high-dimensional generation such as video synthesis.Additional persistent hurdles include maintaining spatial-temporal consistency across complex scenes and mitigating ethical considerations surrounding bias and the preservation of human creative autonomy.This research underscores the transformative potential and inherent complexities of AI-driven synergy within the creative industries.We posit that future research should be dedicated to the synergistic fusion of diffusion and autoregressive models,the integration of multimodal inputs,and the balanced consideration of ethical implications,particularly regarding bias and the preservation of human creative autonomy,thereby establishing a robust foundation for the advancement of anime creation and the broader landscape of AI-driven content generation.展开更多
Sodium-ion batteries have gradually been commercialized due to their wide range of material sources and low cost.However,there are few studies focusing on the commercial sodium-ion battery safety,especially the relati...Sodium-ion batteries have gradually been commercialized due to their wide range of material sources and low cost.However,there are few studies focusing on the commercial sodium-ion battery safety,especially the relationship between heat and gas generation is unclear.This work conducts the thermal runaway(TR)experiments of commercial 18650 sodium-ion batteries with different states of charge(SOCs)under adiabatic accelerated rate calorimetry and localized overheating.The results show that heat generation values of 50% and 100%SOC batteries during TR are 175.2 and 328.2 J g^(-1),respectively.Whereas,0%SOC batteries do not trigger TR.Moreover,the reaction sources and pathways of gas generation during TR are critically sorted out.Finally,two important conclusions are obtained.(i)During the five stages of TR,the heat generation from the safe venting to the triggering of TR stage is the highest in 50%SOC batteries,accounting for 62.5% of the total heat generation.However,for 100%SOC batteries,the heat generation from triggering TR to maximum temperature stage has the largest proportion during TR,at 57%.The 50%SOC batteries present characteristic of slow heat generation,while the 100%SOC batteries show characteristics of accelerated heat generation.(ii)Based on dimensionless analysis,the heat/gas generation ratios of 50% and 100%SOC batteries are 0.262 and 0.028,respectively.The gas generation behavior occur earlier than heat generation behavior during the whole process of TR of sodium-ion batteries.This study provides a direction for the development of high-safety sodium-ion batteries and thermal runaway suppression technology.展开更多
Conventional approaches for obtaining the second and third harmonics typically employ several nonlinear crystals to generate them,which is restricted in application due to the complexity of the optical path and the bu...Conventional approaches for obtaining the second and third harmonics typically employ several nonlinear crystals to generate them,which is restricted in application due to the complexity of the optical path and the bulkiness of the device.In this work,we present a comprehensive theoretical and numerical investigation of the simultaneous generation and competition between the second harmonic waves(SHW)and the third harmonic waves(THW)in a single nonlinear crystal.Through analyzing both small-signal and large-signal regimes,we reveal the complex coupling mechanisms between SHW and THW generation processes.Using periodically poled lithium niobate as an example,we demonstrate that the relative conversion efficiencies between SHW and THW can be freely adjusted by controlling the input fundamental wave power.This work provides new insights for designing efficient frequency converters capable of generating both SHW and THW outputs with controllable intensity ratios.展开更多
Artificial intelligence(AI)assisted ultrasound report generation represents a technology that leverages artificial intelligence to convert ultrasound imaging analysis results into structured diagnostic reports.By inte...Artificial intelligence(AI)assisted ultrasound report generation represents a technology that leverages artificial intelligence to convert ultrasound imaging analysis results into structured diagnostic reports.By integrating image recognition and natural language generation models,AI systems can automatically detect and analyze lesions or abnormalities in ultrasound images,generating textual descriptions of diagnostic conclusions(e.g.,fatty liver,liver fibrosis,automated BIRADS grading of breast lesions),imaging findings,and clinical recommendations to form comprehensive reports.This technology enhances the efficiency and accuracy of imaging diagnosis,reduces physicians’workloads,ensures report standardization and consistency,and provides robust support for clinical decisionmaking.Current state-of-the-art algorithms for automated ultrasound report generation primarily rely on vision-language models,which harness the generalization capabilities of large language models and large vision models through multimodal(language+vision)feature alignment.However,existing approaches inadequately address challenges such as numerical measurement generation,effective utilization of report templates,incorporation of historical reports,learning text-image correlations,and overfitting under limited data conditions.This paper aims to introduce the current state of research on ultrasound report generation,the existing issues,and to provide some thoughts for future research.展开更多
The coal-bearing source rocks in the Jurassic Shuixigou Group have received widespread attention as the primary source rocks in the Turpan-Hami Basin of China,but the hydrocarbon generation potential and process of th...The coal-bearing source rocks in the Jurassic Shuixigou Group have received widespread attention as the primary source rocks in the Turpan-Hami Basin of China,but the hydrocarbon generation potential and process of the mudstone in the Shuixigou Group,especially the mudstone at the top of the Sangonghe Formation,are unclear.Taking the source rocks of the Xishanyao Formation and the Sangonghe Formation as objectives,this study conducted rock pyrolysis and gold tube simulation experiment to investigate their hydrocarbon generation characteristics and differences.Our results indicate that the source rocks of the Xishanyao Formation include mudstone,carbonaceous mudstone and coal,and the quality of the source rocks is highly heterogeneous;the source rocks of the Sangonghe Formation are mainly composed of mudstone,and it is a good gas source rock.Simulation experiments found that the activation energy required for the generation of gaseous hydrocarbons by the mudstone of the Sangonghe Formation is lower than that by the mudstone of the Xishanyao Formation.The hydrocarbon generation process can be divided into three stages for both formations,but the gas generation potential of the Xishanyao Formation mudstone is higher than that of the Sangonghe Formation mudstone.A large amount of hydrocarbon was generated by the mudstone of the Xishanyao Formation when entering late thermal evolution,of which methane is dominant,mainly from the demethylation reaction of mature kerogen.On the other hand,a large amount of hydrocarbon was generated by the mudstone of the Sangonghe Formation in the early stage of thermal evolution,of which light hydrocarbon and wet gas are dominant,mainly from the early cracking stage of kerogen.This difference may be attributed to the structure of kerogen.The mudstone of the Xishanyao Formation is conducive to the formation of highly mature dry gas reservoirs,while the mudstone of the Sangonghe Formation is conducive to the formation of low maturity condensate gas and volatile oil reservoirs.The research result provides a scientific basis for the comparison of oil and gas sources and the evaluation of oil and gas resources in the Turpan-Hami Basin.展开更多
Accurate prediction of the composition of pyrolysis products is the prerequisite for achieving directional regulation of organic-rich shale pyrolysis and conversion products.In this paper,the classical segmented pyrol...Accurate prediction of the composition of pyrolysis products is the prerequisite for achieving directional regulation of organic-rich shale pyrolysis and conversion products.In this paper,the classical segmented pyrolysis kinetics model and a new refined pyrolysis kinetics model were used to forecast the composition distribution of hydrocarbon generation products co-heated by supercritical water and medium and low maturity organic-rich shale.The prediction accuracy of the two reaction kinetics models for the composition of pyrolysis products of organic-rich shale was compared.The reaction path of hydrocarbon generation in centimeter sized organic-rich shale under the action of supercritical water was identified.The results show that the prediction accuracy of the classical segmented pyrolysis kinetics model was poor at the initial stage of the reaction,and gradually increased with increasing time.The prediction error can reach less than 25%when the reaction time was 12 h.The new refined model of reaction kinetics established is better than the classical reaction kinetics model in predicting the product distribution of pyrolysis oil and gas,and its prediction error is less than 14%in this paper.The reaction paths of hydrocarbon generation in centimeter sized organic-rich shale under supercritical water conversion mainly include organic-rich shale directly generates asphaltene and saturated hydrocarbon,asphaltene pyrolysis generates saturated hydrocarbon,aromatic hydrocarbon and resin,saturated hydrocarbon,aromatic hydrocarbon and resin polymerization generates asphaltene,and saturated hydrocarbon,resin and asphaltene generates gas.The reason for the difference of centimeter sized and millimeter sized medium and low maturity organic-rich shales hydrocarbon generation in supercritical water is that the increase of shale size promotes the reaction path of polymerization of saturated hydrocarbon and aromatic hydrocarbon to asphaltene.展开更多
Waveform generation and digitization play essential roles in numerous physics experiments.In traditional distributed systems for large-scale experiments,each frontend node contains an FPGA for data preprocessing,which...Waveform generation and digitization play essential roles in numerous physics experiments.In traditional distributed systems for large-scale experiments,each frontend node contains an FPGA for data preprocessing,which interfaces with various data converters and exchanges data with a backend central processor.However,the streaming readout architecture has become a new paradigm for several experiments benefiting from advancements in data transmission and computing technologies.This paper proposes a scalable distributed waveform generation and digitization system that utilizes fiber optical connections for data transmission between frontend nodes and a central processor.By utilizing transparent transmission on top of the data link layer,the clock and data ports of the converters in the frontend nodes are directly mapped to the FPGA firmware at the backend.This streaming readout architecture reduces the complexity of frontend development and maintains the data conversion in proximity to the detector.Each frontend node uses a local clock for waveform digitization.To translate the timing information of events in each channel into the system clock domain within the backend central processing FPGA,a novel method is proposed and evaluated using a demonstrator system.展开更多
Rice(Oryza sativa L.)plays a pivotal role in global food security,yet its breeding is constrained by its long generation time and seasonality.To enhance rice breeding efficiency and meet future food demands,we have de...Rice(Oryza sativa L.)plays a pivotal role in global food security,yet its breeding is constrained by its long generation time and seasonality.To enhance rice breeding efficiency and meet future food demands,we have developed a vertical hydroponic breeding system integrated with light-emitting diodes(LEDs)light-ing in a closed plant factory(PF),which significantly accelerates rice growth and generation advance-ment.The results show that indica rice can be harvested as early as after 63 days of cultivation,a 50%reduction compared with field cultivation,enabling the annual harvesting of 5-6 generations within the PF.A hyperspectral imaging(HSI)system and attenuated total reflectance infrared(ATR-IR)spec-troscopy were further employed to characterize the chemical composition of the PF-and field-cultivated rice.Metabolomics analysis with ultra-performance liquid chromatography-tandem mass spectrometry(UPLC-MS/MS)and gas chromatography-mass spectrometry(GC-MS)revealed that,com-pared with the field-cultivated rice,the PF-cultivated rice exhibited an up-regulation of total phenolic acids along with 68 non-volatile and 19 volatile metabolites,such as isovitexin,succinic acid,and methylillicinone F.Overall,this study reveals the unique metabolic profile of PF-cultivated rice and high-lights the potential of PFs to accelerate the breeding of crops such as rice,offering an innovative agricul-tural strategy to support food security in the face of global population growth and climate change.展开更多
As a form of discrete representation learning,Vector Quantized Variational Autoencoders(VQ-VAE)have increasingly been applied to generative and multimodal tasks due to their ease of embedding and representative capaci...As a form of discrete representation learning,Vector Quantized Variational Autoencoders(VQ-VAE)have increasingly been applied to generative and multimodal tasks due to their ease of embedding and representative capacity.However,existing VQ-VAEs often perform quantization in the spatial domain,ignoring global structural information and potentially suffering from codebook collapse and information coupling issues.This paper proposes a frequency quantized variational autoencoder(FQ-VAE)to address these issues.The proposed method transforms image features into linear combinations in the frequency domain using a 2D fast Fourier transform(2D-FFT)and performs adaptive quantization on these frequency components to preserve image’s global relationships.The codebook is dynamically optimized to avoid collapse and information coupling issue by considering the usage frequency and dependency of code vectors.Furthermore,we introduce a post-processing module based on graph convolutional networks to further improve reconstruction quality.Experimental results on four public datasets demonstrate that the proposed method outperforms state-of-the-art approaches in terms of Structural Similarity Index(SSIM),Learned Perceptual Image Patch Similarity(LPIPS),and Reconstruction Fréchet Inception Distance(rFID).In the experiments on the CIFAR-10 dataset,compared to the baselinemethod VQ-VAE,the proposedmethod improves the abovemetrics by 4.9%,36.4%,and 52.8%,respectively.展开更多
In the context of the diversity of smart terminals,the unity of the root of trust becomes complicated,which not only affects the efficiency of trust propagation,but also poses a challenge to the security of the whole ...In the context of the diversity of smart terminals,the unity of the root of trust becomes complicated,which not only affects the efficiency of trust propagation,but also poses a challenge to the security of the whole system.In particular,the solidification of the root of trust in non-volatile memory(NVM)restricts the system’s dynamic updating capability,which is an obvious disadvantage in a rapidly changing security environment.To address this issue,this study proposes a novel approach to generate root security parameters using static random access memory(SRAM)physical unclonable functions(PUFs).SRAM PUFs,as a security primitive,show great potential in lightweight security solutions due to their inherent physical properties,low cost and scalability.However,the stability of SRAM PUFs in harsh environments is a key issue.These environmental conditions include extreme temperatures,high humidity,and strong electromagnetic radiation,all of which can affect the performance of SRAM PUFs.In order to ensure the stability of root safety parameters under these conditions,this study proposes an integrated approach that covers not only the acquisition of entropy sources,but also the implementation of algorithms and configuration management.In addition,this study develops a series of reliability-enhancing algorithms,including adaptive parameter selection,data preprocessing,auxiliary data generation,and error correction,which are essential for improving the performance of SRAM PUFs in harsh environments.Based on these techniques,this study establishes six types of secure parameter generation mechanisms,which not only improve the security of the system,but also enhance its adaptability in variable environments.Through a series of experiments,we verify the effectiveness of the proposed method.Under 10 different environmental conditions,our method is able to achieve full recovery of security data with an error rate of less than 25%,which proves the robustness and reliability of our method.These results not only provide strong evidence for the stability of SRAM PUFs in practical applications,but also provide a new direction for future research in the field of smart terminal security.展开更多
Second harmonic generation(SHG),a fundamental and widely-studied phenomenon in nonlinear optics,has attracted significant attention for its ability to convert fundamental frequencies into their second harmonics.While ...Second harmonic generation(SHG),a fundamental and widely-studied phenomenon in nonlinear optics,has attracted significant attention for its ability to convert fundamental frequencies into their second harmonics.While the dominant SHG research has been focused on the optical and infrared regimes,its investigation in the microwave range presents challenges due to the requirements of materials with higher nonlinear coefficients and high-power microwave sources.Here,we provide an overview of methods together with underlying mechanisms for SHG in microwave frequencies,and discuss prospects and insights into the future developments of SHG-based technologies.The discussions on both numerical analyses and experimental studies will offer guidance for further SHG research and communication advancements in microwave regime.展开更多
Ultrasound neuromodulation is a powerful tool for brain investigation and holds great promise for treating brain diseases.However,due to the heterogeneous acoustic properties of skulls,existing ultrasound neuromodulat...Ultrasound neuromodulation is a powerful tool for brain investigation and holds great promise for treating brain diseases.However,due to the heterogeneous acoustic properties of skulls,existing ultrasound neuromodulation faces the challenge of severe transcranial acoustic attenuation.To overcome such limitations,we report an implantable bio-chip for visible and controllable mi-crowave-induced transcranial acoustic generation(MI-tAG).The bio-chip is soft,flexible,and biocompatible,with a thickness of 3mm,making it suitable for human intracranial implantation.The constituted fluid channels can cover an area of 50 mm×60 mm,enabling widefield neuronstimulation.The particles filled in the fluid channels have both high microwave absorption.ensuring efficient ultrasound generation,and magnetism,allowing noncontact and flexible ma-nipulation by external magnetic fields.The experimental results demonstrate that the optimal MI-tAG can be realized by the combination of particles arranged in a linear pattern and corre-sponding illumination via a linearly polarized microwave.Stability evaluation indicates that the particles can maintain a consistent acoustic intensity without degradation for at least seven days.The results of in vitro and in vivo experiments show that the MII-tAG can manipulate ultrasound sources and visibly locate them in real time.This study provides a potential innovative approach for future ultrasound neuromodulation,inspiring the development of more useful methods to advance brain research.This study introduces a promising innovative approach for transcranial acoustic generation,potentially inspiring the development of more effective methods for ad-vancing ultrasound neuromodulation.展开更多
Many existing immune detection algorithms rely on a large volume of labeled self-training samples,which are often difficult to obtain in practical scenarios,thus limiting the training of detection models.Furthermore,n...Many existing immune detection algorithms rely on a large volume of labeled self-training samples,which are often difficult to obtain in practical scenarios,thus limiting the training of detection models.Furthermore,noise inherent in the samples can substantially degrade the detection accuracy of these algorithms.To overcome these challenges,we propose an immune generation algorithm that leverages clustering and a rebound mechanism for label propagation(LP-CRI).The dataset is randomly partitioned into multiple subsets,each of which undergoes clustering followed by label propagation and evaluation.The rebound mechanism assesses the model’s performance after propagation and determines whether to revert to its previous state,initiating a subsequent round of propagation to ensure stable and effective training.Experimental results demonstrate that the proposed method is both computationally efficient and easy to train,significantly enhancing detector performance and outperforming traditional immune detection algorithms.展开更多
The task of molecule generation guided by specific text descriptions has been proposed to generate molecules that match given text inputs.Mainstream methods typically use simplified molecular input line entry system(S...The task of molecule generation guided by specific text descriptions has been proposed to generate molecules that match given text inputs.Mainstream methods typically use simplified molecular input line entry system(SMILES)to represent molecules and rely on diffusion models or autoregressive structures for modeling.However,the one-to-many mapping diversity when using SMILES to represent molecules causes existing methods to require complex model architectures and larger training datasets to improve performance,which affects the efficiency of model training and generation.In this paper,we propose a text-guided diverse-expression diffusion(TGDD)model for molecule generation.TGDD combines both SMILES and self-referencing embedded strings(SELFIES)into a novel diverse-expression molecular representation,enabling precise molecule mapping based on natural language.By leveraging this diverse-expression representation,TGDD simplifies the segmented diffusion generation process,achieving faster training and reduced memory consumption,while also exhibiting stronger alignment with natural language.TGDD outperforms both TGM-LDM and the autoregressive model MolT5-Base on most evaluation metrics.展开更多
Knowledge graphs convey precise semantic information that can be effectively interpreted by neural networks,and generating descriptive text based on these graphs places significant emphasis on content consistency.Howe...Knowledge graphs convey precise semantic information that can be effectively interpreted by neural networks,and generating descriptive text based on these graphs places significant emphasis on content consistency.However,knowledge graphs are inadequate for providing additional linguistic features such as paragraph structure and expressive modes,making it challenging to ensure content coherence in generating text that spans multiple sentences.This lack of coherence can further compromise the overall consistency of the content within a paragraph.In this work,we present the generation of scientific abstracts by leveraging knowledge graphs,with a focus on enhancing both content consistency and coherence.In particular,we construct the ACL Abstract Graph Dataset(ACL-AGD)which pairs knowledge graphs with text,incorporating sentence labels to guide text structure and diverse expressions.We then implement a Siamese network to complement and concretize the entities and relations based on paragraph structure by accomplishing two tasks:graph-to-text generation and entity alignment.Extensive experiments demonstrate that the logical paragraphs generated by our method exhibit entities with a uniform position distribution and appropriate frequency.In terms of content,our method accurately represents the information encoded in the knowledge graph,prevents the generation of irrelevant content,and achieves coherent and non-redundant adjacent sentences,even with a shared knowledge graph.展开更多
Collagen characterization is crucial for disease diagnostics,prevention,and understanding,with growing focus on quantitative analysis at tissue and fibril levels.Numerous models have been developed to quantify structu...Collagen characterization is crucial for disease diagnostics,prevention,and understanding,with growing focus on quantitative analysis at tissue and fibril levels.Numerous models have been developed to quantify structural changes in collagen linked to various pathologies.However,many approaches remain limited to conceptual descriptions or rely on custom software,often requiring programming skills,which re-stricts their clinical application and potential impact.We introduce CollagenFitJ,a plugin for the open-source software platform ImageJ/FIJI,which represents a widely used microscopy image analysis tool.CollagenFitJ makes use of the cylindrical symmetry model for collagen to enable facile quantitative assessment of polarization-resolved second harmonic generation microscopy image stacks.The plugin’s main outputs are collagen structure-related maps(e.g.,orientation and anisotropy of collagen fibrils within the focal volume),which can be accompanied by distribution and randomness maps for a series of structure-related parameters.We describe and validate the use of CollagenFitJ on images acquired on rat-tail tendons,collagen capsules surrounding human thyroid nodules,and mouse colon tumors,using both scanning and widefield second harmonic generation microscopy datasets.The plugin was designed to be user-friendly,requiring little to no experience in image processing and coding to facilitate access for life scientists,medical staff,and microscopy practitioners with limited coding skills or time availability required for coding.展开更多
Research on supercontinuum sources on silicon has made significant progress in the past few decades.However,conventional approaches to broaden the spectral bandwidth often rely on complex and critical dispersion engin...Research on supercontinuum sources on silicon has made significant progress in the past few decades.However,conventional approaches to broaden the spectral bandwidth often rely on complex and critical dispersion engineering by optimizing the core thickness or introducing the cladding with special materials and structures.We propose and demonstrate supercontinuum generation using long-periodgrating(LPG)waveguides on silicon with a C-band pump.The LPG waveguide is introduced for quasi-phase matching,and the generated supercontinuum spectrum is improved greatly with grating-induced dispersive waves.In addition,the demonstrated LPG waveguide shows a low propagation loss comparable with regular silicon photonic waveguides without gratings.In experiments,when using a 1550-nm 75-fs pulse pump with a pulse energy of 200 pJ,the supercontinuum spectrum generated with the present LPG waveguide shows an ultrabroad extent from 1150 to 2300 nm,which is much wider by 200 nm than that achieved by dispersionengineered uniform silicon photonic waveguides on the same chip.This provides a promising option for on-chip broadband light source for silicon photonic systems.展开更多
基金funding from King Saud University through Researchers Supporting Project number(RSP2024R387),King Saud University,Riyadh,Saudi Arabia.
文摘The emergence of next generation networks(NextG),including 5G and beyond,is reshaping the technological landscape of cellular and mobile networks.These networks are sufficiently scaled to interconnect billions of users and devices.Researchers in academia and industry are focusing on technological advancements to achieve highspeed transmission,cell planning,and latency reduction to facilitate emerging applications such as virtual reality,the metaverse,smart cities,smart health,and autonomous vehicles.NextG continuously improves its network functionality to support these applications.Multiple input multiple output(MIMO)technology offers spectral efficiency,dependability,and overall performance in conjunctionwithNextG.This article proposes a secure channel estimation technique in MIMO topology using a norm-estimation model to provide comprehensive insights into protecting NextG network components against adversarial attacks.The technique aims to create long-lasting and secure NextG networks using this extended approach.The viability of MIMO applications and modern AI-driven methodologies to combat cybersecurity threats are explored in this research.Moreover,the proposed model demonstrates high performance in terms of reliability and accuracy,with a 20%reduction in the MalOut-RealOut-Diff metric compared to existing state-of-the-art techniques.
文摘To investigate the overall performance of reverse energy bypass scramjet,firstly a variable spe⁃cific heat method combined with a chemical balance calculation module for combustion products were used to es⁃tablish a benchmark scramjet performance evaluation model.Based on the test data of typical flying point of Mach 7 with the altitude of 29 km,the reliability of the model was verified.The deviations of parameters such as the to⁃tal pressure loss of combustor between the model and the test data were analyzed.Furtherly,an analytical method for post-combustion magnetohydrodynamic power generation was established;by embedding the above method into the overall performance evaluation model,performance prediction considering the power generation effect was realized.Finally,based on the above model,variety regulations of the inlet and the outlet parameters of the power generation channel and performance parameters including the engine specific impulse and the unit thrust under different enthalpy extraction ratios and load factors were analyzed.It could be concluded that the model can reliably predict the variations of key parameters.As the value of the load factor increases,the value of the conduc⁃tivity required to reach the specified enthalpy extraction ratio first decreases and then increases,which is approxi⁃mately parabolic.In order to reduce the demand for the gas conductivity for MHD power generation,the load fac⁃tor should be around 0.5.When the load factor is 0.4 and the magnetic induction intensity is 2.5 T,if the enthalpy extraction ratio reaches 0.5%,the engine specific impulse performance reduces about 3.58%.
基金supported by the National Nat-ural Science Foundation of China(Nos.12192251,12174185,92163216,and 62288101).
文摘The generation of optical vortices from nonlinear photonic crystals(NPCs)with spatially modulated second-order nonlinearity offers a promising approach to extend the working wavelength and topological charge of vortex beams for various applications.In this work,the second harmonic(SH)optical vortex beams generated from nonlinear fork gratings under Gaussian beam illumination are numerically investigated.The far-field intensity and phase distributions,as well as the orbital angular momentum(OAM)spectra of the SH beams,are analyzed for different structural topological charges and diffraction orders.Results reveal that higher-order diffraction and larger structural topological charges lead to angular interference patterns and non-uniform intensity distributions,deviating from the standard vortex profile.To optimize the SH vortex quality,the effects of the fundamental wave beam waist,crystal thickness,and grating duty cycle are explored.It is shown that increasing the beam waist can effectively suppress diffraction order interference and improve the beam’s quality.This study provides theoretical guidance for enhancing the performance of nonlinear optical devices based on NPCs.
基金supported by the National Natural Science Foundation of China(Grant No.62202210).
文摘The application of generative artificial intelligence(AI)is bringing about notable changes in anime creation.This paper surveys recent advancements and applications of diffusion and language models in anime generation,focusing on their demonstrated potential to enhance production efficiency through automation and personalization.Despite these benefits,it is crucial to acknowledge the substantial initial computational investments required for training and deploying these models.We conduct an in-depth survey of cutting-edge generative AI technologies,encompassing models such as Stable Diffusion and GPT,and appraise pivotal large-scale datasets alongside quantifiable evaluation metrics.Review of the surveyed literature indicates the achievement of considerable maturity in the capacity of AI models to synthesize high-quality,aesthetically compelling anime visual images from textual prompts,alongside discernible progress in the generation of coherent narratives.However,achieving perfect long-form consistency,mitigating artifacts like flickering in video sequences,and enabling fine-grained artistic control remain critical ongoing challenges.Building upon these advancements,research efforts have increasingly pivoted towards the synthesis of higher-dimensional content,such as video and three-dimensional assets,with recent studies demonstrating significant progress in this burgeoning field.Nevertheless,formidable challenges endure amidst these advancements.Foremost among these are the substantial computational exigencies requisite for training and deploying these sophisticated models,particularly pronounced in the realm of high-dimensional generation such as video synthesis.Additional persistent hurdles include maintaining spatial-temporal consistency across complex scenes and mitigating ethical considerations surrounding bias and the preservation of human creative autonomy.This research underscores the transformative potential and inherent complexities of AI-driven synergy within the creative industries.We posit that future research should be dedicated to the synergistic fusion of diffusion and autoregressive models,the integration of multimodal inputs,and the balanced consideration of ethical implications,particularly regarding bias and the preservation of human creative autonomy,thereby establishing a robust foundation for the advancement of anime creation and the broader landscape of AI-driven content generation.
基金supported by the National Key R&D Program of China(No.2024YFE0209200)National Natural Science Foundation of China(No.52404259)+1 种基金Postgraduate Academic Innovation Project of Anhui Province(No.2023xscx009)supported by Youth Innovation Promotion Association CAS(No.Y201768)。
文摘Sodium-ion batteries have gradually been commercialized due to their wide range of material sources and low cost.However,there are few studies focusing on the commercial sodium-ion battery safety,especially the relationship between heat and gas generation is unclear.This work conducts the thermal runaway(TR)experiments of commercial 18650 sodium-ion batteries with different states of charge(SOCs)under adiabatic accelerated rate calorimetry and localized overheating.The results show that heat generation values of 50% and 100%SOC batteries during TR are 175.2 and 328.2 J g^(-1),respectively.Whereas,0%SOC batteries do not trigger TR.Moreover,the reaction sources and pathways of gas generation during TR are critically sorted out.Finally,two important conclusions are obtained.(i)During the five stages of TR,the heat generation from the safe venting to the triggering of TR stage is the highest in 50%SOC batteries,accounting for 62.5% of the total heat generation.However,for 100%SOC batteries,the heat generation from triggering TR to maximum temperature stage has the largest proportion during TR,at 57%.The 50%SOC batteries present characteristic of slow heat generation,while the 100%SOC batteries show characteristics of accelerated heat generation.(ii)Based on dimensionless analysis,the heat/gas generation ratios of 50% and 100%SOC batteries are 0.262 and 0.028,respectively.The gas generation behavior occur earlier than heat generation behavior during the whole process of TR of sodium-ion batteries.This study provides a direction for the development of high-safety sodium-ion batteries and thermal runaway suppression technology.
基金supported by the Science and Technology Project of Guangdong Province,China(Grant No.2020B010190001)the National Natural Science Foundation of China(Grant No.12434016)+1 种基金the National Key Research and Development Program of China(Grant No.2023YFA1406900)the Fund of the National Postdoctoral Researcher Program(Grant No.GZB20240785).
文摘Conventional approaches for obtaining the second and third harmonics typically employ several nonlinear crystals to generate them,which is restricted in application due to the complexity of the optical path and the bulkiness of the device.In this work,we present a comprehensive theoretical and numerical investigation of the simultaneous generation and competition between the second harmonic waves(SHW)and the third harmonic waves(THW)in a single nonlinear crystal.Through analyzing both small-signal and large-signal regimes,we reveal the complex coupling mechanisms between SHW and THW generation processes.Using periodically poled lithium niobate as an example,we demonstrate that the relative conversion efficiencies between SHW and THW can be freely adjusted by controlling the input fundamental wave power.This work provides new insights for designing efficient frequency converters capable of generating both SHW and THW outputs with controllable intensity ratios.
文摘Artificial intelligence(AI)assisted ultrasound report generation represents a technology that leverages artificial intelligence to convert ultrasound imaging analysis results into structured diagnostic reports.By integrating image recognition and natural language generation models,AI systems can automatically detect and analyze lesions or abnormalities in ultrasound images,generating textual descriptions of diagnostic conclusions(e.g.,fatty liver,liver fibrosis,automated BIRADS grading of breast lesions),imaging findings,and clinical recommendations to form comprehensive reports.This technology enhances the efficiency and accuracy of imaging diagnosis,reduces physicians’workloads,ensures report standardization and consistency,and provides robust support for clinical decisionmaking.Current state-of-the-art algorithms for automated ultrasound report generation primarily rely on vision-language models,which harness the generalization capabilities of large language models and large vision models through multimodal(language+vision)feature alignment.However,existing approaches inadequately address challenges such as numerical measurement generation,effective utilization of report templates,incorporation of historical reports,learning text-image correlations,and overfitting under limited data conditions.This paper aims to introduce the current state of research on ultrasound report generation,the existing issues,and to provide some thoughts for future research.
基金supported by the China Petroleum Science and Technology Major Project(No.2023ZZ18-03).
文摘The coal-bearing source rocks in the Jurassic Shuixigou Group have received widespread attention as the primary source rocks in the Turpan-Hami Basin of China,but the hydrocarbon generation potential and process of the mudstone in the Shuixigou Group,especially the mudstone at the top of the Sangonghe Formation,are unclear.Taking the source rocks of the Xishanyao Formation and the Sangonghe Formation as objectives,this study conducted rock pyrolysis and gold tube simulation experiment to investigate their hydrocarbon generation characteristics and differences.Our results indicate that the source rocks of the Xishanyao Formation include mudstone,carbonaceous mudstone and coal,and the quality of the source rocks is highly heterogeneous;the source rocks of the Sangonghe Formation are mainly composed of mudstone,and it is a good gas source rock.Simulation experiments found that the activation energy required for the generation of gaseous hydrocarbons by the mudstone of the Sangonghe Formation is lower than that by the mudstone of the Xishanyao Formation.The hydrocarbon generation process can be divided into three stages for both formations,but the gas generation potential of the Xishanyao Formation mudstone is higher than that of the Sangonghe Formation mudstone.A large amount of hydrocarbon was generated by the mudstone of the Xishanyao Formation when entering late thermal evolution,of which methane is dominant,mainly from the demethylation reaction of mature kerogen.On the other hand,a large amount of hydrocarbon was generated by the mudstone of the Sangonghe Formation in the early stage of thermal evolution,of which light hydrocarbon and wet gas are dominant,mainly from the early cracking stage of kerogen.This difference may be attributed to the structure of kerogen.The mudstone of the Xishanyao Formation is conducive to the formation of highly mature dry gas reservoirs,while the mudstone of the Sangonghe Formation is conducive to the formation of low maturity condensate gas and volatile oil reservoirs.The research result provides a scientific basis for the comparison of oil and gas sources and the evaluation of oil and gas resources in the Turpan-Hami Basin.
基金support by the Basic Science Center Program of the Ordered Energy Conversion of the National Nature Science Foundation of China(NO.52488201)is gratefully acknowledged.
文摘Accurate prediction of the composition of pyrolysis products is the prerequisite for achieving directional regulation of organic-rich shale pyrolysis and conversion products.In this paper,the classical segmented pyrolysis kinetics model and a new refined pyrolysis kinetics model were used to forecast the composition distribution of hydrocarbon generation products co-heated by supercritical water and medium and low maturity organic-rich shale.The prediction accuracy of the two reaction kinetics models for the composition of pyrolysis products of organic-rich shale was compared.The reaction path of hydrocarbon generation in centimeter sized organic-rich shale under the action of supercritical water was identified.The results show that the prediction accuracy of the classical segmented pyrolysis kinetics model was poor at the initial stage of the reaction,and gradually increased with increasing time.The prediction error can reach less than 25%when the reaction time was 12 h.The new refined model of reaction kinetics established is better than the classical reaction kinetics model in predicting the product distribution of pyrolysis oil and gas,and its prediction error is less than 14%in this paper.The reaction paths of hydrocarbon generation in centimeter sized organic-rich shale under supercritical water conversion mainly include organic-rich shale directly generates asphaltene and saturated hydrocarbon,asphaltene pyrolysis generates saturated hydrocarbon,aromatic hydrocarbon and resin,saturated hydrocarbon,aromatic hydrocarbon and resin polymerization generates asphaltene,and saturated hydrocarbon,resin and asphaltene generates gas.The reason for the difference of centimeter sized and millimeter sized medium and low maturity organic-rich shales hydrocarbon generation in supercritical water is that the increase of shale size promotes the reaction path of polymerization of saturated hydrocarbon and aromatic hydrocarbon to asphaltene.
基金supported by the National Key Research and Development Program of China(No.2022YFA1604703)the National Natural Science Foundation of China(No.12375189)the National Key Research and Development Program of China(No.2021YFA1601300)。
文摘Waveform generation and digitization play essential roles in numerous physics experiments.In traditional distributed systems for large-scale experiments,each frontend node contains an FPGA for data preprocessing,which interfaces with various data converters and exchanges data with a backend central processor.However,the streaming readout architecture has become a new paradigm for several experiments benefiting from advancements in data transmission and computing technologies.This paper proposes a scalable distributed waveform generation and digitization system that utilizes fiber optical connections for data transmission between frontend nodes and a central processor.By utilizing transparent transmission on top of the data link layer,the clock and data ports of the converters in the frontend nodes are directly mapped to the FPGA firmware at the backend.This streaming readout architecture reduces the complexity of frontend development and maintains the data conversion in proximity to the detector.Each frontend node uses a local clock for waveform digitization.To translate the timing information of events in each channel into the system clock domain within the backend central processing FPGA,a novel method is proposed and evaluated using a demonstrator system.
基金supported by the National Key Research and Development Program(2023YFF1001500)the Local Financial Funds of National Agricultural Science and Technology Center,Chengdu(NASC2022KR02,NASC2023TD08,NASC2021ST08,NASC2021PC04,NASC2022KR07,NASC2022KR06,and NASC2023ST04)+2 种基金the Agricultural Science and Technology Innova-tion Program(ASTIP-34-IUA-01,ASTIP-34-IUA-02,ASTIP-IUA-2023003,and ASTIP2024-34-IUA-09)the Central Public-interest Scientific Institution Basal Research Fund(Y2023YJ07 and SZ202403)the Sichuan Science and Technology Program(2023YFN003,2024NSFC1261,2023YFQ0100,and 2023ZYD0089).
文摘Rice(Oryza sativa L.)plays a pivotal role in global food security,yet its breeding is constrained by its long generation time and seasonality.To enhance rice breeding efficiency and meet future food demands,we have developed a vertical hydroponic breeding system integrated with light-emitting diodes(LEDs)light-ing in a closed plant factory(PF),which significantly accelerates rice growth and generation advance-ment.The results show that indica rice can be harvested as early as after 63 days of cultivation,a 50%reduction compared with field cultivation,enabling the annual harvesting of 5-6 generations within the PF.A hyperspectral imaging(HSI)system and attenuated total reflectance infrared(ATR-IR)spec-troscopy were further employed to characterize the chemical composition of the PF-and field-cultivated rice.Metabolomics analysis with ultra-performance liquid chromatography-tandem mass spectrometry(UPLC-MS/MS)and gas chromatography-mass spectrometry(GC-MS)revealed that,com-pared with the field-cultivated rice,the PF-cultivated rice exhibited an up-regulation of total phenolic acids along with 68 non-volatile and 19 volatile metabolites,such as isovitexin,succinic acid,and methylillicinone F.Overall,this study reveals the unique metabolic profile of PF-cultivated rice and high-lights the potential of PFs to accelerate the breeding of crops such as rice,offering an innovative agricul-tural strategy to support food security in the face of global population growth and climate change.
基金supported by the Interdisciplinary project of Dalian University DLUXK-2023-ZD-001.
文摘As a form of discrete representation learning,Vector Quantized Variational Autoencoders(VQ-VAE)have increasingly been applied to generative and multimodal tasks due to their ease of embedding and representative capacity.However,existing VQ-VAEs often perform quantization in the spatial domain,ignoring global structural information and potentially suffering from codebook collapse and information coupling issues.This paper proposes a frequency quantized variational autoencoder(FQ-VAE)to address these issues.The proposed method transforms image features into linear combinations in the frequency domain using a 2D fast Fourier transform(2D-FFT)and performs adaptive quantization on these frequency components to preserve image’s global relationships.The codebook is dynamically optimized to avoid collapse and information coupling issue by considering the usage frequency and dependency of code vectors.Furthermore,we introduce a post-processing module based on graph convolutional networks to further improve reconstruction quality.Experimental results on four public datasets demonstrate that the proposed method outperforms state-of-the-art approaches in terms of Structural Similarity Index(SSIM),Learned Perceptual Image Patch Similarity(LPIPS),and Reconstruction Fréchet Inception Distance(rFID).In the experiments on the CIFAR-10 dataset,compared to the baselinemethod VQ-VAE,the proposedmethod improves the abovemetrics by 4.9%,36.4%,and 52.8%,respectively.
基金supported by National key Research and Development Program“Security Protection Technology for Critical Information Infrastructure of Distribution Network”(2022YFB3105100).
文摘In the context of the diversity of smart terminals,the unity of the root of trust becomes complicated,which not only affects the efficiency of trust propagation,but also poses a challenge to the security of the whole system.In particular,the solidification of the root of trust in non-volatile memory(NVM)restricts the system’s dynamic updating capability,which is an obvious disadvantage in a rapidly changing security environment.To address this issue,this study proposes a novel approach to generate root security parameters using static random access memory(SRAM)physical unclonable functions(PUFs).SRAM PUFs,as a security primitive,show great potential in lightweight security solutions due to their inherent physical properties,low cost and scalability.However,the stability of SRAM PUFs in harsh environments is a key issue.These environmental conditions include extreme temperatures,high humidity,and strong electromagnetic radiation,all of which can affect the performance of SRAM PUFs.In order to ensure the stability of root safety parameters under these conditions,this study proposes an integrated approach that covers not only the acquisition of entropy sources,but also the implementation of algorithms and configuration management.In addition,this study develops a series of reliability-enhancing algorithms,including adaptive parameter selection,data preprocessing,auxiliary data generation,and error correction,which are essential for improving the performance of SRAM PUFs in harsh environments.Based on these techniques,this study establishes six types of secure parameter generation mechanisms,which not only improve the security of the system,but also enhance its adaptability in variable environments.Through a series of experiments,we verify the effectiveness of the proposed method.Under 10 different environmental conditions,our method is able to achieve full recovery of security data with an error rate of less than 25%,which proves the robustness and reliability of our method.These results not only provide strong evidence for the stability of SRAM PUFs in practical applications,but also provide a new direction for future research in the field of smart terminal security.
基金supported by the National Natural Science Foun-dation of China(No.12274339).
文摘Second harmonic generation(SHG),a fundamental and widely-studied phenomenon in nonlinear optics,has attracted significant attention for its ability to convert fundamental frequencies into their second harmonics.While the dominant SHG research has been focused on the optical and infrared regimes,its investigation in the microwave range presents challenges due to the requirements of materials with higher nonlinear coefficients and high-power microwave sources.Here,we provide an overview of methods together with underlying mechanisms for SHG in microwave frequencies,and discuss prospects and insights into the future developments of SHG-based technologies.The discussions on both numerical analyses and experimental studies will offer guidance for further SHG research and communication advancements in microwave regime.
基金supported by the National Key R&D Program of China under grant 2023YFF0715303in part by the National Natural Science Foundation of China under Grant Nos.62305148,62105140,62022037,and 61775028+2 种基金in part by the Department of Science and Technology of Guangdong Province under Grant Nos.2019ZT08Y191 and 2022B1212010003in part by the Shenzhen Science and Technology Program under Grant Nos.JCYJ20220530114010023,RCJC20231211090039066,20231116104616001,KQTD20190929172743294,JCYJ20230807093105010,RCBS20231211090802011in part by the Startup Grant from Southern University of Science and Technology under Grant No.PDJH2021C008.
文摘Ultrasound neuromodulation is a powerful tool for brain investigation and holds great promise for treating brain diseases.However,due to the heterogeneous acoustic properties of skulls,existing ultrasound neuromodulation faces the challenge of severe transcranial acoustic attenuation.To overcome such limitations,we report an implantable bio-chip for visible and controllable mi-crowave-induced transcranial acoustic generation(MI-tAG).The bio-chip is soft,flexible,and biocompatible,with a thickness of 3mm,making it suitable for human intracranial implantation.The constituted fluid channels can cover an area of 50 mm×60 mm,enabling widefield neuronstimulation.The particles filled in the fluid channels have both high microwave absorption.ensuring efficient ultrasound generation,and magnetism,allowing noncontact and flexible ma-nipulation by external magnetic fields.The experimental results demonstrate that the optimal MI-tAG can be realized by the combination of particles arranged in a linear pattern and corre-sponding illumination via a linearly polarized microwave.Stability evaluation indicates that the particles can maintain a consistent acoustic intensity without degradation for at least seven days.The results of in vitro and in vivo experiments show that the MII-tAG can manipulate ultrasound sources and visibly locate them in real time.This study provides a potential innovative approach for future ultrasound neuromodulation,inspiring the development of more useful methods to advance brain research.This study introduces a promising innovative approach for transcranial acoustic generation,potentially inspiring the development of more effective methods for ad-vancing ultrasound neuromodulation.
基金granted by Key Project of Beijing Municipal Social Science Foundation(No.15ZHA004)Key Project of Beijing Municipal Social Science Foundation and Beijing Municipal Education Commission Social Science Program(No.SZ20231123202).
文摘Many existing immune detection algorithms rely on a large volume of labeled self-training samples,which are often difficult to obtain in practical scenarios,thus limiting the training of detection models.Furthermore,noise inherent in the samples can substantially degrade the detection accuracy of these algorithms.To overcome these challenges,we propose an immune generation algorithm that leverages clustering and a rebound mechanism for label propagation(LP-CRI).The dataset is randomly partitioned into multiple subsets,each of which undergoes clustering followed by label propagation and evaluation.The rebound mechanism assesses the model’s performance after propagation and determines whether to revert to its previous state,initiating a subsequent round of propagation to ensure stable and effective training.Experimental results demonstrate that the proposed method is both computationally efficient and easy to train,significantly enhancing detector performance and outperforming traditional immune detection algorithms.
基金supported in part by the National Natural Science Foundation of China(Grant Nos.62476247 and 62072409)the“Pioneer”and“Leading Goose”R&D Program of Zhejiang(Grant No.2024C01214)the Zhejiang Provincial Natural Science Foundation(Grant No.LR21F020003).
文摘The task of molecule generation guided by specific text descriptions has been proposed to generate molecules that match given text inputs.Mainstream methods typically use simplified molecular input line entry system(SMILES)to represent molecules and rely on diffusion models or autoregressive structures for modeling.However,the one-to-many mapping diversity when using SMILES to represent molecules causes existing methods to require complex model architectures and larger training datasets to improve performance,which affects the efficiency of model training and generation.In this paper,we propose a text-guided diverse-expression diffusion(TGDD)model for molecule generation.TGDD combines both SMILES and self-referencing embedded strings(SELFIES)into a novel diverse-expression molecular representation,enabling precise molecule mapping based on natural language.By leveraging this diverse-expression representation,TGDD simplifies the segmented diffusion generation process,achieving faster training and reduced memory consumption,while also exhibiting stronger alignment with natural language.TGDD outperforms both TGM-LDM and the autoregressive model MolT5-Base on most evaluation metrics.
文摘Knowledge graphs convey precise semantic information that can be effectively interpreted by neural networks,and generating descriptive text based on these graphs places significant emphasis on content consistency.However,knowledge graphs are inadequate for providing additional linguistic features such as paragraph structure and expressive modes,making it challenging to ensure content coherence in generating text that spans multiple sentences.This lack of coherence can further compromise the overall consistency of the content within a paragraph.In this work,we present the generation of scientific abstracts by leveraging knowledge graphs,with a focus on enhancing both content consistency and coherence.In particular,we construct the ACL Abstract Graph Dataset(ACL-AGD)which pairs knowledge graphs with text,incorporating sentence labels to guide text structure and diverse expressions.We then implement a Siamese network to complement and concretize the entities and relations based on paragraph structure by accomplishing two tasks:graph-to-text generation and entity alignment.Extensive experiments demonstrate that the logical paragraphs generated by our method exhibit entities with a uniform position distribution and appropriate frequency.In terms of content,our method accurately represents the information encoded in the knowledge graph,prevents the generation of irrelevant content,and achieves coherent and non-redundant adjacent sentences,even with a shared knowledge graph.
基金supported by the Ministry of Research,Innovation and Digitalization,CNCS-UEFISCDI[Grant Nos.RO-NO-2019-0601(MEDYCONAI),PN-III-P4-PCE-2021-0444(RESONANO)]PN-IV-P1-PCE-2023-1137+2 种基金supported in part by IN2SIGHT,European Union’s Horizon 2020(GA.no.964481)by the Research Council of Lithuania(LMTLTAgreement No.P-MIP-23-237).
文摘Collagen characterization is crucial for disease diagnostics,prevention,and understanding,with growing focus on quantitative analysis at tissue and fibril levels.Numerous models have been developed to quantify structural changes in collagen linked to various pathologies.However,many approaches remain limited to conceptual descriptions or rely on custom software,often requiring programming skills,which re-stricts their clinical application and potential impact.We introduce CollagenFitJ,a plugin for the open-source software platform ImageJ/FIJI,which represents a widely used microscopy image analysis tool.CollagenFitJ makes use of the cylindrical symmetry model for collagen to enable facile quantitative assessment of polarization-resolved second harmonic generation microscopy image stacks.The plugin’s main outputs are collagen structure-related maps(e.g.,orientation and anisotropy of collagen fibrils within the focal volume),which can be accompanied by distribution and randomness maps for a series of structure-related parameters.We describe and validate the use of CollagenFitJ on images acquired on rat-tail tendons,collagen capsules surrounding human thyroid nodules,and mouse colon tumors,using both scanning and widefield second harmonic generation microscopy datasets.The plugin was designed to be user-friendly,requiring little to no experience in image processing and coding to facilitate access for life scientists,medical staff,and microscopy practitioners with limited coding skills or time availability required for coding.
基金supported by the UK’s Engineering and Physical Sciences Research Council(Grant Nos.EP/V000624/1,EP/X03495X/1,EP/X041166/1,and EP/T02643X/1)the Royal Society(Grant No.RG\R2\232531).
文摘Research on supercontinuum sources on silicon has made significant progress in the past few decades.However,conventional approaches to broaden the spectral bandwidth often rely on complex and critical dispersion engineering by optimizing the core thickness or introducing the cladding with special materials and structures.We propose and demonstrate supercontinuum generation using long-periodgrating(LPG)waveguides on silicon with a C-band pump.The LPG waveguide is introduced for quasi-phase matching,and the generated supercontinuum spectrum is improved greatly with grating-induced dispersive waves.In addition,the demonstrated LPG waveguide shows a low propagation loss comparable with regular silicon photonic waveguides without gratings.In experiments,when using a 1550-nm 75-fs pulse pump with a pulse energy of 200 pJ,the supercontinuum spectrum generated with the present LPG waveguide shows an ultrabroad extent from 1150 to 2300 nm,which is much wider by 200 nm than that achieved by dispersionengineered uniform silicon photonic waveguides on the same chip.This provides a promising option for on-chip broadband light source for silicon photonic systems.
