X-ray detectors show potential applications in medical imaging,materials science,and nuclear energy.To achieve high detection efficiency and spatial resolution,many conventional semiconductor materials,such as amorpho...X-ray detectors show potential applications in medical imaging,materials science,and nuclear energy.To achieve high detection efficiency and spatial resolution,many conventional semiconductor materials,such as amorphous selenium,cadmium telluride zinc,and perovskites,have been utilized in direct conversion X-ray detectors.However,these semiconductor materials are susceptible to temperature-induced performance degradation,crystallization,delamination,uneven lattice growth,radiation damage,and high dark current.This study explores a new approach by coupling an FC40 electronic fluorinated liquid with a specialized high-resolution and high-readout-speed complementary metal-oxide-semiconductor(CMOS)pixel array,specifically the Topmetal II−chip,to fabricate a direct conversion X-ray detector.The fluorinated liquid FC40(molecular formula:C_(21)F_(48)N_(2))is an electronic medium that is minimally affected by temperature and displays no issues with uniform conductivity.It exhibits a low dark current and minimal radiation damage and enables customizable thickness in X-ray absorption.This addresses the limitations inherent in conventional semiconductor-based detectors.In this study,simple X-ray detector imaging tests were conducted,demonstrating the excellent coupling capability between FC40 electronic fluorinated liquid and CMOS chips by the X-ray detector.A spatial resolution of 4.0 lp/mm was measured using a striped line par card,and a relatively clear image of a cockroach was displayed in the digital radiography imaging results.Preliminary test results indicated the feasibility of fabricating an X-ray detector by combining FC40 electronic fluorinated liquid and CMOS chips.Owing to the absence of issues related to chip-material coupling,a high spatial resolution could be achieved by reducing the chip pixel size.This method presents a new avenue for studies on novel liquid-based direct conversion X-ray detectors.展开更多
The PICOSEC Micromegas(MM)is a precise timing gaseous detector based on a Cherenkov radiator coupled with a semi-transparent photocathode and an MM amplifying structure.It features a two-stage amplification process th...The PICOSEC Micromegas(MM)is a precise timing gaseous detector based on a Cherenkov radiator coupled with a semi-transparent photocathode and an MM amplifying structure.It features a two-stage amplification process that leads to a significant deterioration of non-uniformity when scaling up to larger areas.Since the performance of gaseous detectors is highly dependent on the choice of working gas,optimizing the gas mixture offers a promising solution to improve the uniformity performance.This paper addresses these challenges through a combined approach of simulation based on Garfield++and experimental studies.The simulation investigates the properties of different mixing fractions of gas mixtures and their impact on detector performance,including gain uniformity and time resolution.To verify the simulation results,experimental tests were conducted using a multi-channel PICOSEC MM prototype with different gas mixtures.The experimental results are consistent with the findings of the simulation,indicating that a higher concentration of neon significantly improves the detector’s gain uniformity.Furthermore,the influence of gas mixtures on time resolution was explored as a critical performance indicator.The study presented in this paper offers valuable insights for improving uniformity in large-area PICOSEC MM detectors and optimizing overall performance.展开更多
To detect space gravitational waves in the extremely low-frequency band,the telescope and optic-al platform require high stability and reliability.However,the cantilevered design presents challenges,espe-cially in the...To detect space gravitational waves in the extremely low-frequency band,the telescope and optic-al platform require high stability and reliability.However,the cantilevered design presents challenges,espe-cially in the glass-metal hetero-bonding process.This study focuses on the analysis and experimental re-search of the bonding layer in the integrated structure.By optimizing the structural configuration and select-ing suitable bonding processes,the reliability of the telescope system is enhanced.The research indicates that using J-133 adhesive achieves the best performance,with a bonding layer thickness of 0.30 mm and a metal substrate surface roughness of Ra 0.8.These findings significantly enhance the reliability of the optical sys-tem while minimizing potential risks.展开更多
The study of the charge conjugation and parity(CP)violation of hyperon is the precision frontier for probing possible new CP violation sources beyond the standard model(SM).With the large number of quantum entangled h...The study of the charge conjugation and parity(CP)violation of hyperon is the precision frontier for probing possible new CP violation sources beyond the standard model(SM).With the large number of quantum entangled hyperonantihyperon pairs to be produced at Super Tau-Charm Facility(STCF),the CP asymmetry of hyperon is expected to be tested with a statistical sensitivity of 10^(−4) or even better.To cope with the statistical precision,the systematic effects from various aspects are critical and need to be studied in detail.In this paper,the sensitivity effects on the CP violation parameters associated with the detector resolution,including those of the position and momentum,are studied and discussed in detail.The results provide valuable guidance for the design of STCF detector.展开更多
For segmented detectors,surface flatness is critical as it directly influences both energy resolution and image clarity.Additionally,the limited adjustment range of the segmented detectors necessitates precise benchma...For segmented detectors,surface flatness is critical as it directly influences both energy resolution and image clarity.Additionally,the limited adjustment range of the segmented detectors necessitates precise benchmark construction.This paper proposes an architecture for detecting detector flatness based on channel spectral dispersion.By measuring the dispersion fringes for coplanar adjustment,the final adjustment residual is improved to better than 300 nm.This result validates the feasibility of the proposed technology and provides significant technical support for the development of next-generation large-aperture sky survey equipment.展开更多
Radiation doses to patients in diagnostics and interventional radiology need to be optimized to comply with the principles of radiation protection in medical practice. This involves using specific detectors with respe...Radiation doses to patients in diagnostics and interventional radiology need to be optimized to comply with the principles of radiation protection in medical practice. This involves using specific detectors with respective diagnostic beams to carry out quality control/quality assurance tests needed to optimize patient doses in the hospital. Semiconductor detectors are used in dosimetry to verify the equipment performance and dose to patients. This work aims to assess the performance, energy dependence, and response of five commercially available semiconductor detectors in RQR, RQR-M, RQA, and RQT at Secondary Standard Dosimetry for clinical applications. The diagnostic beams were generated using Exradin A4 reference ion chamber and PTW electrometer. The ambient temperature and pressure were noted for KTP correction. The detectors designed for RQR showed good performance in RQT beams and vice versa. The detectors designed for RQR-M displayed high energy dependency in other diagnostic beams. The type of diagnostic beam quality determines the response of semiconductor detectors. Therefore, a detector should be calibrated according to the beam qualities to be measured.展开更多
The polarization properties of light are widely applied in imaging,communications,materials analy⁃sis,and life sciences.Various methods have been developed that can measure the polarization information of a target.How...The polarization properties of light are widely applied in imaging,communications,materials analy⁃sis,and life sciences.Various methods have been developed that can measure the polarization information of a target.However,conventional polarization detection systems are often bulky and complex,limiting their poten⁃tial for broader applications.To address the challenges of miniaturization,integrated polarization detectors have been extensively explored in recent years,achieving significant advancements in performance and functionality.In this review,we focus mainly on integrated polarization detectors with innovative features,including infinitely high polarization discrimination,ultrahigh sensitivity to polarization state change,full Stokes parameters measure⁃ment,and simultaneous perception of polarization and other key properties of light.Lastly,we discuss the oppor⁃tunities and challenges for the future development of integrated polarization photodetectors.展开更多
Sarcasm detection is a complex and challenging task,particularly in the context of Chinese social media,where it exhibits strong contextual dependencies and cultural specificity.To address the limitations of existing ...Sarcasm detection is a complex and challenging task,particularly in the context of Chinese social media,where it exhibits strong contextual dependencies and cultural specificity.To address the limitations of existing methods in capturing the implicit semantics and contextual associations in sarcastic expressions,this paper proposes an event-aware model for Chinese sarcasm detection,leveraging a multi-head attention(MHA)mechanism and contrastive learning(CL)strategies.The proposed model employs a dual-path Bidirectional Encoder Representations from Transformers(BERT)encoder to process comment text and event context separately and integrates an MHA mechanism to facilitate deep interactions between the two,thereby capturing multidimensional semantic associations.Additionally,a CL strategy is introduced to enhance feature representation capabilities,further improving the model’s performance in handling class imbalance and complex contextual scenarios.The model achieves state-of-the-art performance on the Chinese sarcasm dataset,with significant improvements in accuracy(79.55%),F1-score(84.22%),and an area under the curve(AUC,84.35%).展开更多
Position-sensitive neutron detectors play an important role in neutron scattering studies. Detectors based on ~6LiF/ZnS(Ag) scintillator and wave-shifting fiber have the advantages of high neutron detection efficiency...Position-sensitive neutron detectors play an important role in neutron scattering studies. Detectors based on ~6LiF/ZnS(Ag) scintillator and wave-shifting fiber have the advantages of high neutron detection efficiency, high position resolution,and large-area splicing, and can well meet the requirement of large area neutron detection for neutron diffractometers. An engineering detector prototype based on a ~6LiF/ZnS(Ag) scintillation screen and SiPM array readout was fabricated for the General Purpose Powder Diffractometer of China Spallation Neutron Source(CSNS). The detector has an active area of 196 mm × 444 mm, with a pixel size of 4 mm × 4 mm. The key performances of the detector prototype were tested at the BL20 neutron beam line of CSNS. The test results show that the neutron detection efficiency of the detector was 32% and 42% at wavelengths of 1.4 ? and 2.8 ?, respectively. An interpolated neutron detection efficiency of 40.2% at a wavelength of 2 ? was obtained. The tested neutron efficiency non-uniformity of the detector was 10.2%, which is less than one-half that of the current general purpose powder diffractometer scintillator neutron detectors at CSNS. This work achieves, for the first time, an efficiency uniformity of < 11% in large-area mosaic neutron detectors, alongside significant advancements in electromagnetic interference immunity and cost-effectiveness.展开更多
As the group-buying model shows significant progress in attracting new users,enhancing user engagement,and increasing platform profitability,providing personalized recommendations for group-buying users has emerged as...As the group-buying model shows significant progress in attracting new users,enhancing user engagement,and increasing platform profitability,providing personalized recommendations for group-buying users has emerged as a new challenge in the field of recommendation systems.This paper introduces a group-buying recommendation model based on multi-head attention mechanisms and multi-task learning,termed the Multi-head Attention Mechanisms and Multi-task Learning Group-Buying Recommendation(MAMGBR)model,specifically designed to optimize group-buying recommendations on e-commerce platforms.The core dataset of this study comes from the Chinese maternal and infant e-commerce platform“Beibei,”encompassing approximately 430,000 successful groupbuying actions and over 120,000 users.Themodel focuses on twomain tasks:recommending items for group organizers(Task Ⅰ)and recommending participants for a given group-buying event(Task Ⅱ).In model evaluation,MAMGBR achieves an MRR@10 of 0.7696 for Task I,marking a 20.23%improvement over baseline models.Furthermore,in Task II,where complex interaction patterns prevail,MAMGBR utilizes auxiliary loss functions to effectively model the multifaceted roles of users,items,and participants,leading to a 24.08%increase in MRR@100 under a 1:99 sample ratio.Experimental results show that compared to benchmark models,such as NGCF and EATNN,MAMGBR’s integration ofmulti-head attentionmechanisms,expert networks,and gating mechanisms enables more accurate modeling of user preferences and social associations within group-buying scenarios,significantly enhancing recommendation accuracy and platform group-buying success rates.展开更多
Safety maintenance of power equipment is of great importance in power grids,in which image-processing-based defect recognition is supposed to classify abnormal conditions during daily inspection.However,owing to the b...Safety maintenance of power equipment is of great importance in power grids,in which image-processing-based defect recognition is supposed to classify abnormal conditions during daily inspection.However,owing to the blurred features of defect images,the current defect recognition algorithm has poor fine-grained recognition ability.Visual attention can achieve fine-grained recognition with its abil-ity to model long-range dependencies while introducing extra computational complexity,especially for multi-head attention in vision transformer structures.Under these circumstances,this paper proposes a self-reduction multi-head attention module that can reduce computational complexity and be easily combined with a Convolutional Neural Network(CNN).In this manner,local and global fea-tures can be calculated simultaneously in our proposed structure,aiming to improve the defect recognition performance.Specifically,the proposed self-reduction multi-head attention can reduce redundant parameters,thereby solving the problem of limited computational resources.Experimental results were obtained based on the defect dataset collected from the substation.The results demonstrated the efficiency and superiority of the proposed method over other advanced algorithms.展开更多
The measurement of low-level radioactivity using high-purity germanium(HPGe)detectors is important in applications such as environmental background radiation,material screening,and rare decays.The dead layers,dead zon...The measurement of low-level radioactivity using high-purity germanium(HPGe)detectors is important in applications such as environmental background radiation,material screening,and rare decays.The dead layers,dead zones,aluminum shell thickness,and diameter of Ge crystals are the most influential factors affecting the performance of HPGe detectors;hence,precise modeling of the physical conditions of the detectors is highly desirable.In this study,the GEANT4 simulation framework with an optimized detector geometry adequately replicated the experimentally recorded spectrum.These detector simulations explored the idea of realizing a dead zone(an inactive volume)at the backend of an n-type coaxial Gecrystal.Using multigamma sources,the effect of true coincidence summing(TCS)on the full energy peak(FEP)efficiency calibration of an HPGe detector was investigated as a function of sample-to-detector distance.Good agreements between the simulated and experimental efficiencies as well as the simulated and analytically calculated summing coincidence correction coefficients were achieved.At a short distance between the source and detector,calculating the correction factors for a strong source posed challenges owing to significant deadtime and pile-up effects of the detection system.The described methodology can efficiently determine summing peak probabilities at short sample-to-detector distances.展开更多
Abnormal network traffic, as a frequent security risk, requires a series of techniques to categorize and detect it. Existing network traffic anomaly detection still faces challenges: the inability to fully extract loc...Abnormal network traffic, as a frequent security risk, requires a series of techniques to categorize and detect it. Existing network traffic anomaly detection still faces challenges: the inability to fully extract local and global features, as well as the lack of effective mechanisms to capture complex interactions between features;Additionally, when increasing the receptive field to obtain deeper feature representations, the reliance on increasing network depth leads to a significant increase in computational resource consumption, affecting the efficiency and performance of detection. Based on these issues, firstly, this paper proposes a network traffic anomaly detection model based on parallel dilated convolution and residual learning (Res-PDC). To better explore the interactive relationships between features, the traffic samples are converted into two-dimensional matrix. A module combining parallel dilated convolutions and residual learning (res-pdc) was designed to extract local and global features of traffic at different scales. By utilizing res-pdc modules with different dilation rates, we can effectively capture spatial features at different scales and explore feature dependencies spanning wider regions without increasing computational resources. Secondly, to focus and integrate the information in different feature subspaces, further enhance and extract the interactions among the features, multi-head attention is added to Res-PDC, resulting in the final model: multi-head attention enhanced parallel dilated convolution and residual learning (MHA-Res-PDC) for network traffic anomaly detection. Finally, comparisons with other machine learning and deep learning algorithms are conducted on the NSL-KDD and CIC-IDS-2018 datasets. The experimental results demonstrate that the proposed method in this paper can effectively improve the detection performance.展开更多
A state-of-the-art detector array with a digital data acquisition system has been developed for charged-particle decay studies,includingβ-delayed protons,αdecay,and direct proton emissions from exotic proton-rich nu...A state-of-the-art detector array with a digital data acquisition system has been developed for charged-particle decay studies,includingβ-delayed protons,αdecay,and direct proton emissions from exotic proton-rich nuclei.The digital data acquisition system enables precise synchronization and processing of complex signals from various detectors,such as plastic scintillators,silicon detectors,and germaniumγdetectors.The system's performance was evaluated using theβdecay of^(32)Ar and its neighboring nuclei,produced via projectile fragmentation at the first Radioactive Ion Beam Line in Lanzhou(RIBLL1).Key measurements,including the half-life,charged-particle spectrum,andγ-ray spectrum,were obtained and compared with previous results for validation.Using the implantation–decay method,the isotopes of interest were implanted into two doublesided silicon strip detectors,where their subsequent decays were measured and correlated with preceding implantations using both position and time information.This detection system has potential for further applications,including the study ofβ-delayed charged-particle decay and direct proton emissions from even more exotic proton-rich nuclei.展开更多
Lightweight deep learning models are increasingly required in resource-constrained environments such as mobile devices and the Internet of Medical Things(IoMT).Multi-head convolution with channel attention can facilit...Lightweight deep learning models are increasingly required in resource-constrained environments such as mobile devices and the Internet of Medical Things(IoMT).Multi-head convolution with channel attention can facilitate learning activations relevant to different kernel sizes within a multi-head convolutional layer.Therefore,this study investigates the capability of novel lightweight models incorporating residual multi-head convolution with channel attention(ResMHCNN)blocks to classify medical images.We introduced three novel lightweight deep learning models(BT-Net,LCC-Net,and BC-Net)utilizing the ResMHCNN block as their backbone.These models were crossvalidated and tested on three publicly available medical image datasets:a brain tumor dataset from Figshare consisting of T1-weighted magnetic resonance imaging slices of meningioma,glioma,and pituitary tumors;the LC25000 dataset,which includes microscopic images of lung and colon cancers;and the BreaKHis dataset,containing benign and malignant breast microscopic images.The lightweight models achieved accuracies of 96.9%for 3-class brain tumor classification using BT-Net,and 99.7%for 5-class lung and colon cancer classification using LCC-Net.For 2-class breast cancer classification,BC-Net achieved an accuracy of 96.7%.The parameter counts for the proposed lightweight models—LCC-Net,BC-Net,and BT-Net—are 0.528,0.226,and 1.154 million,respectively.The presented lightweight models,featuring ResMHCNN blocks,may be effectively employed for accurate medical image classification.In the future,these models might be tested for viability in resource-constrained systems such as mobile devices and IoMT platforms.展开更多
Coal dust explosions are severe safety accidents in coal mine production,posing significant threats to life and property.Predicting the maximum explosion pressure(Pm)of coal dust using deep learning models can effecti...Coal dust explosions are severe safety accidents in coal mine production,posing significant threats to life and property.Predicting the maximum explosion pressure(Pm)of coal dust using deep learning models can effectively assess potential risks and provide a scientific basis for preventing coal dust explosions.In this study,a 20-L explosion sphere apparatus was used to test the maximum explosion pressure of coal dust under seven different particle sizes and ten mass concentrations(Cdust),resulting in a dataset of 70 experimental groups.Through Spearman correlation analysis and random forest feature selection methods,particle size(D_(10),D_(20),D_(50))and mass concentration(Cdust)were identified as critical feature parameters from the ten initial parameters of the coal dust samples.Based on this,a hybrid Long Short-Term Memory(LSTM)network model incorporating a Multi-Head Attention Mechanism and the Sparrow Search Algorithm(SSA)was proposed to predict the maximum explosion pressure of coal dust.The results demonstrate that the SSA-LSTM-Multi-Head Attention model excels in predicting the maximum explosion pressure of coal dust.The four evaluation metrics indicate that the model achieved a coefficient of determination(R^(2)),root mean square error(RMSE),mean absolute percentage error(MAPE),and mean absolute error(MAE)of 0.9841,0.0030,0.0074,and 0.0049,respectively,in the training set.In the testing set,these values were 0.9743,0.0087,0.0108,and 0.0069,respectively.Compared to artificial neural networks(ANN),random forest(RF),support vector machines(SVM),particle swarm optimized-SVM(PSO-SVM)neural networks,and the traditional single-model LSTM,the SSA-LSTM-Multi-Head Attention model demonstrated superior generalization capability and prediction accuracy.The findings of this study not only advance the application of deep learning in coal dust explosion prediction but also provide robust technical support for the prevention and risk assessment of coal dust explosions.展开更多
Superconducting kinetic inductance detectors(KIDs)are considered to be a highly promising technique for the large-scale imaging of millimeter and submillimeter waves in astronomy.As the pixel density and the array siz...Superconducting kinetic inductance detectors(KIDs)are considered to be a highly promising technique for the large-scale imaging of millimeter and submillimeter waves in astronomy.As the pixel density and the array size increase,the electromagnetic crosstalk inevitably becomes a problem that prevents increasing the multiplexing during the development of larger KIDs arrays.In this work,an effective method is introduced to suppress the electromagnetic crosstalk and achieve a compact pixel distribution and small frequency intervals.The electromagnetic crosstalk is first analyzed by simulating the behavior of two neighboring pixels,and the physical distance and the frequency interval are optimized.Then,the arrangement of the pixels on the whole array is redesigned using a genetic algorithm to satisfy the requirements.The simulation results reveal that the normalized electromagnetic crosstalk can be reduced to 0.5%on an 8×8 array.Larger arrays of 16×16 pixels have been fabricated and measured to validate this method,and the results reveal that both the resonance property and survival rate of pixels are improved effectively with this method.This method will be very helpful for designing high-multiplexing KIDs arrays within a limited bandwidth.展开更多
The self-attention mechanism of Transformers,which captures long-range contextual information,has demonstrated significant potential in image segmentation.However,their ability to learn local,contextual relationships ...The self-attention mechanism of Transformers,which captures long-range contextual information,has demonstrated significant potential in image segmentation.However,their ability to learn local,contextual relationships between pixels requires further improvement.Previous methods face challenges in efficiently managing multi-scale fea-tures of different granularities from the encoder backbone,leaving room for improvement in their global representation and feature extraction capabilities.To address these challenges,we propose a novel Decoder with Multi-Head Feature Receptors(DMHFR),which receives multi-scale features from the encoder backbone and organizes them into three feature groups with different granularities:coarse,fine-grained,and full set.These groups are subsequently processed by Multi-Head Feature Receptors(MHFRs)after feature capture and modeling operations.MHFRs include two Three-Head Feature Receptors(THFRs)and one Four-Head Feature Receptor(FHFR).Each group of features is passed through these MHFRs and then fed into axial transformers,which help the model capture long-range dependencies within the features.The three MHFRs produce three distinct feature outputs.The output from the FHFR serves as auxiliary auxiliary features in the prediction head,and the prediction output and their losses will eventually be aggregated.Experimental results show that the Transformer using DMHFR outperforms 15 state of the arts(SOTA)methods on five public datasets.Specifically,it achieved significant improvements in mean DICE scores over the classic Parallel Reverse Attention Network(PraNet)method,with gains of 4.1%,2.2%,1.4%,8.9%,and 16.3%on the CVC-ClinicDB,Kvasir-SEG,CVC-T,CVC-ColonDB,and ETIS-LaribPolypDB datasets,respectively.展开更多
The prompt fission neutron spectrum(PFNS)is a key nuclear data quantity that is of particular interest and plays a crucial role in understanding and modeling fission processes.An array comprising 48 liquid scintillati...The prompt fission neutron spectrum(PFNS)is a key nuclear data quantity that is of particular interest and plays a crucial role in understanding and modeling fission processes.An array comprising 48 liquid scintillation detectors and a parallelplate avalanche counter(PPAC)was developed at the China Institute of Atomic Energy(CIAE)to measure the PFNS of actinide nuclei.Efficiency and energy calibrations were performed for all the liquid scintillators,and their efficiencies were consistently found to be better than 5%.The time resolutions of the PPAC and liquid scintillators were measured to be 1.08 ns and 1.16 ns using~(252)Cf and~(207)Bi sources,respectively.The pulse shape discrimination of the liquid scintillator was utilized to identify neutron andγsignals on an event-by-event basis,and the figure of merit was deduced as 1.12 at a 200 ke Vee threshold.The contribution to the PFNS from multiple scattered neutrons was evaluated via Geant4 simulations,and those originating from the environment were found to be comparable to the crosstalk between the detectors.The neutron efficiency of the entire detection array was calibrated using a~(252)Cf spontaneous fission source and was demonstrated to be consistent with the Geant4 simulation results,which verified the reliability of the detection array.展开更多
基金supported by the National Natural Science Foundation of China(No.12235006)the National Key Research and Development Program of China(No.2020YFE0202002.
文摘X-ray detectors show potential applications in medical imaging,materials science,and nuclear energy.To achieve high detection efficiency and spatial resolution,many conventional semiconductor materials,such as amorphous selenium,cadmium telluride zinc,and perovskites,have been utilized in direct conversion X-ray detectors.However,these semiconductor materials are susceptible to temperature-induced performance degradation,crystallization,delamination,uneven lattice growth,radiation damage,and high dark current.This study explores a new approach by coupling an FC40 electronic fluorinated liquid with a specialized high-resolution and high-readout-speed complementary metal-oxide-semiconductor(CMOS)pixel array,specifically the Topmetal II−chip,to fabricate a direct conversion X-ray detector.The fluorinated liquid FC40(molecular formula:C_(21)F_(48)N_(2))is an electronic medium that is minimally affected by temperature and displays no issues with uniform conductivity.It exhibits a low dark current and minimal radiation damage and enables customizable thickness in X-ray absorption.This addresses the limitations inherent in conventional semiconductor-based detectors.In this study,simple X-ray detector imaging tests were conducted,demonstrating the excellent coupling capability between FC40 electronic fluorinated liquid and CMOS chips by the X-ray detector.A spatial resolution of 4.0 lp/mm was measured using a striped line par card,and a relatively clear image of a cockroach was displayed in the digital radiography imaging results.Preliminary test results indicated the feasibility of fabricating an X-ray detector by combining FC40 electronic fluorinated liquid and CMOS chips.Owing to the absence of issues related to chip-material coupling,a high spatial resolution could be achieved by reducing the chip pixel size.This method presents a new avenue for studies on novel liquid-based direct conversion X-ray detectors.
基金supported by the National Natural Science Foundation of China(12125505).
文摘The PICOSEC Micromegas(MM)is a precise timing gaseous detector based on a Cherenkov radiator coupled with a semi-transparent photocathode and an MM amplifying structure.It features a two-stage amplification process that leads to a significant deterioration of non-uniformity when scaling up to larger areas.Since the performance of gaseous detectors is highly dependent on the choice of working gas,optimizing the gas mixture offers a promising solution to improve the uniformity performance.This paper addresses these challenges through a combined approach of simulation based on Garfield++and experimental studies.The simulation investigates the properties of different mixing fractions of gas mixtures and their impact on detector performance,including gain uniformity and time resolution.To verify the simulation results,experimental tests were conducted using a multi-channel PICOSEC MM prototype with different gas mixtures.The experimental results are consistent with the findings of the simulation,indicating that a higher concentration of neon significantly improves the detector’s gain uniformity.Furthermore,the influence of gas mixtures on time resolution was explored as a critical performance indicator.The study presented in this paper offers valuable insights for improving uniformity in large-area PICOSEC MM detectors and optimizing overall performance.
文摘To detect space gravitational waves in the extremely low-frequency band,the telescope and optic-al platform require high stability and reliability.However,the cantilevered design presents challenges,espe-cially in the glass-metal hetero-bonding process.This study focuses on the analysis and experimental re-search of the bonding layer in the integrated structure.By optimizing the structural configuration and select-ing suitable bonding processes,the reliability of the telescope system is enhanced.The research indicates that using J-133 adhesive achieves the best performance,with a bonding layer thickness of 0.30 mm and a metal substrate surface roughness of Ra 0.8.These findings significantly enhance the reliability of the optical sys-tem while minimizing potential risks.
基金supported by the National Key R&D Program of China(2022YFA1602200)the International Partnership Program of the Chinese Academy of Sciences(211134KYSB20200057).
文摘The study of the charge conjugation and parity(CP)violation of hyperon is the precision frontier for probing possible new CP violation sources beyond the standard model(SM).With the large number of quantum entangled hyperonantihyperon pairs to be produced at Super Tau-Charm Facility(STCF),the CP asymmetry of hyperon is expected to be tested with a statistical sensitivity of 10^(−4) or even better.To cope with the statistical precision,the systematic effects from various aspects are critical and need to be studied in detail.In this paper,the sensitivity effects on the CP violation parameters associated with the detector resolution,including those of the position and momentum,are studied and discussed in detail.The results provide valuable guidance for the design of STCF detector.
文摘For segmented detectors,surface flatness is critical as it directly influences both energy resolution and image clarity.Additionally,the limited adjustment range of the segmented detectors necessitates precise benchmark construction.This paper proposes an architecture for detecting detector flatness based on channel spectral dispersion.By measuring the dispersion fringes for coplanar adjustment,the final adjustment residual is improved to better than 300 nm.This result validates the feasibility of the proposed technology and provides significant technical support for the development of next-generation large-aperture sky survey equipment.
文摘Radiation doses to patients in diagnostics and interventional radiology need to be optimized to comply with the principles of radiation protection in medical practice. This involves using specific detectors with respective diagnostic beams to carry out quality control/quality assurance tests needed to optimize patient doses in the hospital. Semiconductor detectors are used in dosimetry to verify the equipment performance and dose to patients. This work aims to assess the performance, energy dependence, and response of five commercially available semiconductor detectors in RQR, RQR-M, RQA, and RQT at Secondary Standard Dosimetry for clinical applications. The diagnostic beams were generated using Exradin A4 reference ion chamber and PTW electrometer. The ambient temperature and pressure were noted for KTP correction. The detectors designed for RQR showed good performance in RQT beams and vice versa. The detectors designed for RQR-M displayed high energy dependency in other diagnostic beams. The type of diagnostic beam quality determines the response of semiconductor detectors. Therefore, a detector should be calibrated according to the beam qualities to be measured.
基金Supported by the National Key Research and Development Program of China(2022YFA1404602)the Strategic Priority Research Program of the Chinese Academy of Sciences(XDB0580000)+3 种基金the National Natural Science Foundation of China(U23B2045,62305362)the Program of Shanghai Academic/Technology Research Leader(22XD1424400)the Fund of SITP Innovation Foundation(CX-461 and CX-522)Special Project to Seize the Commanding Heights of Science and Technology of Chinese Academy of Sciences,subtopic(GJ0090406-6).
文摘The polarization properties of light are widely applied in imaging,communications,materials analy⁃sis,and life sciences.Various methods have been developed that can measure the polarization information of a target.However,conventional polarization detection systems are often bulky and complex,limiting their poten⁃tial for broader applications.To address the challenges of miniaturization,integrated polarization detectors have been extensively explored in recent years,achieving significant advancements in performance and functionality.In this review,we focus mainly on integrated polarization detectors with innovative features,including infinitely high polarization discrimination,ultrahigh sensitivity to polarization state change,full Stokes parameters measure⁃ment,and simultaneous perception of polarization and other key properties of light.Lastly,we discuss the oppor⁃tunities and challenges for the future development of integrated polarization photodetectors.
基金granted by Qin Xin Talents Cultivation Program(No.QXTCP C202115),Beijing Information Science&Technology Universitythe Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing Fund(No.GJJ-23),National Social Science Foundation,China(No.21BTQ079).
文摘Sarcasm detection is a complex and challenging task,particularly in the context of Chinese social media,where it exhibits strong contextual dependencies and cultural specificity.To address the limitations of existing methods in capturing the implicit semantics and contextual associations in sarcastic expressions,this paper proposes an event-aware model for Chinese sarcasm detection,leveraging a multi-head attention(MHA)mechanism and contrastive learning(CL)strategies.The proposed model employs a dual-path Bidirectional Encoder Representations from Transformers(BERT)encoder to process comment text and event context separately and integrates an MHA mechanism to facilitate deep interactions between the two,thereby capturing multidimensional semantic associations.Additionally,a CL strategy is introduced to enhance feature representation capabilities,further improving the model’s performance in handling class imbalance and complex contextual scenarios.The model achieves state-of-the-art performance on the Chinese sarcasm dataset,with significant improvements in accuracy(79.55%),F1-score(84.22%),and an area under the curve(AUC,84.35%).
基金Project supported by the National Natural Science Foundation of China (Grant No. 12275181)Guangdong Basic and Applied Basic Research Foundation (Grant No. 2022B1515120071)+1 种基金Promotion Project of Scientific Research Capability of Key Construction Disciplines in Guangdong Province (Grant No. 2022ZDJS118)Natural Science Foundation of Top Talent SZTU (Grant No. GDRC202205)。
文摘Position-sensitive neutron detectors play an important role in neutron scattering studies. Detectors based on ~6LiF/ZnS(Ag) scintillator and wave-shifting fiber have the advantages of high neutron detection efficiency, high position resolution,and large-area splicing, and can well meet the requirement of large area neutron detection for neutron diffractometers. An engineering detector prototype based on a ~6LiF/ZnS(Ag) scintillation screen and SiPM array readout was fabricated for the General Purpose Powder Diffractometer of China Spallation Neutron Source(CSNS). The detector has an active area of 196 mm × 444 mm, with a pixel size of 4 mm × 4 mm. The key performances of the detector prototype were tested at the BL20 neutron beam line of CSNS. The test results show that the neutron detection efficiency of the detector was 32% and 42% at wavelengths of 1.4 ? and 2.8 ?, respectively. An interpolated neutron detection efficiency of 40.2% at a wavelength of 2 ? was obtained. The tested neutron efficiency non-uniformity of the detector was 10.2%, which is less than one-half that of the current general purpose powder diffractometer scintillator neutron detectors at CSNS. This work achieves, for the first time, an efficiency uniformity of < 11% in large-area mosaic neutron detectors, alongside significant advancements in electromagnetic interference immunity and cost-effectiveness.
基金supported by the Key Research and Development Program of Heilongjiang Province(No.2022ZX01A35).
文摘As the group-buying model shows significant progress in attracting new users,enhancing user engagement,and increasing platform profitability,providing personalized recommendations for group-buying users has emerged as a new challenge in the field of recommendation systems.This paper introduces a group-buying recommendation model based on multi-head attention mechanisms and multi-task learning,termed the Multi-head Attention Mechanisms and Multi-task Learning Group-Buying Recommendation(MAMGBR)model,specifically designed to optimize group-buying recommendations on e-commerce platforms.The core dataset of this study comes from the Chinese maternal and infant e-commerce platform“Beibei,”encompassing approximately 430,000 successful groupbuying actions and over 120,000 users.Themodel focuses on twomain tasks:recommending items for group organizers(Task Ⅰ)and recommending participants for a given group-buying event(Task Ⅱ).In model evaluation,MAMGBR achieves an MRR@10 of 0.7696 for Task I,marking a 20.23%improvement over baseline models.Furthermore,in Task II,where complex interaction patterns prevail,MAMGBR utilizes auxiliary loss functions to effectively model the multifaceted roles of users,items,and participants,leading to a 24.08%increase in MRR@100 under a 1:99 sample ratio.Experimental results show that compared to benchmark models,such as NGCF and EATNN,MAMGBR’s integration ofmulti-head attentionmechanisms,expert networks,and gating mechanisms enables more accurate modeling of user preferences and social associations within group-buying scenarios,significantly enhancing recommendation accuracy and platform group-buying success rates.
基金supported in part by Major Program of the National Natural Science Foundation of China under Grant 62127803.
文摘Safety maintenance of power equipment is of great importance in power grids,in which image-processing-based defect recognition is supposed to classify abnormal conditions during daily inspection.However,owing to the blurred features of defect images,the current defect recognition algorithm has poor fine-grained recognition ability.Visual attention can achieve fine-grained recognition with its abil-ity to model long-range dependencies while introducing extra computational complexity,especially for multi-head attention in vision transformer structures.Under these circumstances,this paper proposes a self-reduction multi-head attention module that can reduce computational complexity and be easily combined with a Convolutional Neural Network(CNN).In this manner,local and global fea-tures can be calculated simultaneously in our proposed structure,aiming to improve the defect recognition performance.Specifically,the proposed self-reduction multi-head attention can reduce redundant parameters,thereby solving the problem of limited computational resources.Experimental results were obtained based on the defect dataset collected from the substation.The results demonstrated the efficiency and superiority of the proposed method over other advanced algorithms.
基金supported by the Natural Science Foundation of Gansu Province(No.22JR5RA118)the National Natural Science Foundation of China(Nos.12121005 and U1932138)the Strategic Priority Research Program of the Chinese Academy of Sciences(No.XDB34010000)。
文摘The measurement of low-level radioactivity using high-purity germanium(HPGe)detectors is important in applications such as environmental background radiation,material screening,and rare decays.The dead layers,dead zones,aluminum shell thickness,and diameter of Ge crystals are the most influential factors affecting the performance of HPGe detectors;hence,precise modeling of the physical conditions of the detectors is highly desirable.In this study,the GEANT4 simulation framework with an optimized detector geometry adequately replicated the experimentally recorded spectrum.These detector simulations explored the idea of realizing a dead zone(an inactive volume)at the backend of an n-type coaxial Gecrystal.Using multigamma sources,the effect of true coincidence summing(TCS)on the full energy peak(FEP)efficiency calibration of an HPGe detector was investigated as a function of sample-to-detector distance.Good agreements between the simulated and experimental efficiencies as well as the simulated and analytically calculated summing coincidence correction coefficients were achieved.At a short distance between the source and detector,calculating the correction factors for a strong source posed challenges owing to significant deadtime and pile-up effects of the detection system.The described methodology can efficiently determine summing peak probabilities at short sample-to-detector distances.
基金supported by the Xiamen Science and Technology Subsidy Project(No.2023CXY0318).
文摘Abnormal network traffic, as a frequent security risk, requires a series of techniques to categorize and detect it. Existing network traffic anomaly detection still faces challenges: the inability to fully extract local and global features, as well as the lack of effective mechanisms to capture complex interactions between features;Additionally, when increasing the receptive field to obtain deeper feature representations, the reliance on increasing network depth leads to a significant increase in computational resource consumption, affecting the efficiency and performance of detection. Based on these issues, firstly, this paper proposes a network traffic anomaly detection model based on parallel dilated convolution and residual learning (Res-PDC). To better explore the interactive relationships between features, the traffic samples are converted into two-dimensional matrix. A module combining parallel dilated convolutions and residual learning (res-pdc) was designed to extract local and global features of traffic at different scales. By utilizing res-pdc modules with different dilation rates, we can effectively capture spatial features at different scales and explore feature dependencies spanning wider regions without increasing computational resources. Secondly, to focus and integrate the information in different feature subspaces, further enhance and extract the interactions among the features, multi-head attention is added to Res-PDC, resulting in the final model: multi-head attention enhanced parallel dilated convolution and residual learning (MHA-Res-PDC) for network traffic anomaly detection. Finally, comparisons with other machine learning and deep learning algorithms are conducted on the NSL-KDD and CIC-IDS-2018 datasets. The experimental results demonstrate that the proposed method in this paper can effectively improve the detection performance.
基金supported by the National Key Research and Development Project,China(No.2023YFA1606404)the Strategic Priority Research Program of Chinese Academy of Sciences(No.XDB34010300)+5 种基金the National Natural Science Foundation of China(Nos.12022501,12105329,12475127)the Guangdong Major Project of Basic and Applied Basic Research(No.2021B0301030006)the Research Program of Heavy Ion Science and Technology Key Laboratory,Institute of Modern Physics,Chinese Academy of Sciences(Nos.HIST2024KS04,HIST2024CO04)Longyuan Youth Innovation and Entrepreneurship Talent Project of Gansu Province(No.2024GZT04)State Key Laboratory of Nuclear Physics and Technology,Peking University(No.NPT2023KFY01)the Major Science and Technology Projects in Gansu Province(No.24GD13GA005)。
文摘A state-of-the-art detector array with a digital data acquisition system has been developed for charged-particle decay studies,includingβ-delayed protons,αdecay,and direct proton emissions from exotic proton-rich nuclei.The digital data acquisition system enables precise synchronization and processing of complex signals from various detectors,such as plastic scintillators,silicon detectors,and germaniumγdetectors.The system's performance was evaluated using theβdecay of^(32)Ar and its neighboring nuclei,produced via projectile fragmentation at the first Radioactive Ion Beam Line in Lanzhou(RIBLL1).Key measurements,including the half-life,charged-particle spectrum,andγ-ray spectrum,were obtained and compared with previous results for validation.Using the implantation–decay method,the isotopes of interest were implanted into two doublesided silicon strip detectors,where their subsequent decays were measured and correlated with preceding implantations using both position and time information.This detection system has potential for further applications,including the study ofβ-delayed charged-particle decay and direct proton emissions from even more exotic proton-rich nuclei.
基金supported by the Institute of Information&Communications Technology Planning&Evaluation(IITP)-Innovative Human Resource Development for Local Intellectualization program grant funded by the Korea government(MSIT)(IITP-2025-RS-2023-00259678)by INHA UNIVERSITY Research Grant.
文摘Lightweight deep learning models are increasingly required in resource-constrained environments such as mobile devices and the Internet of Medical Things(IoMT).Multi-head convolution with channel attention can facilitate learning activations relevant to different kernel sizes within a multi-head convolutional layer.Therefore,this study investigates the capability of novel lightweight models incorporating residual multi-head convolution with channel attention(ResMHCNN)blocks to classify medical images.We introduced three novel lightweight deep learning models(BT-Net,LCC-Net,and BC-Net)utilizing the ResMHCNN block as their backbone.These models were crossvalidated and tested on three publicly available medical image datasets:a brain tumor dataset from Figshare consisting of T1-weighted magnetic resonance imaging slices of meningioma,glioma,and pituitary tumors;the LC25000 dataset,which includes microscopic images of lung and colon cancers;and the BreaKHis dataset,containing benign and malignant breast microscopic images.The lightweight models achieved accuracies of 96.9%for 3-class brain tumor classification using BT-Net,and 99.7%for 5-class lung and colon cancer classification using LCC-Net.For 2-class breast cancer classification,BC-Net achieved an accuracy of 96.7%.The parameter counts for the proposed lightweight models—LCC-Net,BC-Net,and BT-Net—are 0.528,0.226,and 1.154 million,respectively.The presented lightweight models,featuring ResMHCNN blocks,may be effectively employed for accurate medical image classification.In the future,these models might be tested for viability in resource-constrained systems such as mobile devices and IoMT platforms.
基金funded by the Research on Intelligent Mining Geological Model and Ventilation Model for Extremely Thin Coal Seam in Heilongjiang Province,China(2021ZXJ02A03)the Demonstration of Intelligent Mining for Comprehensive Mining Face in Extremely Thin Coal Seam in Heilongjiang Province,China(2021ZXJ02A04)the Natural Science Foundation of Heilongjiang Province,China(LH2024E112).
文摘Coal dust explosions are severe safety accidents in coal mine production,posing significant threats to life and property.Predicting the maximum explosion pressure(Pm)of coal dust using deep learning models can effectively assess potential risks and provide a scientific basis for preventing coal dust explosions.In this study,a 20-L explosion sphere apparatus was used to test the maximum explosion pressure of coal dust under seven different particle sizes and ten mass concentrations(Cdust),resulting in a dataset of 70 experimental groups.Through Spearman correlation analysis and random forest feature selection methods,particle size(D_(10),D_(20),D_(50))and mass concentration(Cdust)were identified as critical feature parameters from the ten initial parameters of the coal dust samples.Based on this,a hybrid Long Short-Term Memory(LSTM)network model incorporating a Multi-Head Attention Mechanism and the Sparrow Search Algorithm(SSA)was proposed to predict the maximum explosion pressure of coal dust.The results demonstrate that the SSA-LSTM-Multi-Head Attention model excels in predicting the maximum explosion pressure of coal dust.The four evaluation metrics indicate that the model achieved a coefficient of determination(R^(2)),root mean square error(RMSE),mean absolute percentage error(MAPE),and mean absolute error(MAE)of 0.9841,0.0030,0.0074,and 0.0049,respectively,in the training set.In the testing set,these values were 0.9743,0.0087,0.0108,and 0.0069,respectively.Compared to artificial neural networks(ANN),random forest(RF),support vector machines(SVM),particle swarm optimized-SVM(PSO-SVM)neural networks,and the traditional single-model LSTM,the SSA-LSTM-Multi-Head Attention model demonstrated superior generalization capability and prediction accuracy.The findings of this study not only advance the application of deep learning in coal dust explosion prediction but also provide robust technical support for the prevention and risk assessment of coal dust explosions.
基金supported by the National Key Research and Development Program of China(2023YFC2206601)the National Natural Science Foundation of China(12273024,62205211)the Science and Technology Commission of Shanghai Municipality(23010503900,22590780100).
文摘Superconducting kinetic inductance detectors(KIDs)are considered to be a highly promising technique for the large-scale imaging of millimeter and submillimeter waves in astronomy.As the pixel density and the array size increase,the electromagnetic crosstalk inevitably becomes a problem that prevents increasing the multiplexing during the development of larger KIDs arrays.In this work,an effective method is introduced to suppress the electromagnetic crosstalk and achieve a compact pixel distribution and small frequency intervals.The electromagnetic crosstalk is first analyzed by simulating the behavior of two neighboring pixels,and the physical distance and the frequency interval are optimized.Then,the arrangement of the pixels on the whole array is redesigned using a genetic algorithm to satisfy the requirements.The simulation results reveal that the normalized electromagnetic crosstalk can be reduced to 0.5%on an 8×8 array.Larger arrays of 16×16 pixels have been fabricated and measured to validate this method,and the results reveal that both the resonance property and survival rate of pixels are improved effectively with this method.This method will be very helpful for designing high-multiplexing KIDs arrays within a limited bandwidth.
基金supported by Xiamen Medical and Health Guidance Project in 2021(No.3502Z20214ZD1070)supported by a grant from Guangxi Key Laboratory of Machine Vision and Intelligent Control,China(No.2023B02).
文摘The self-attention mechanism of Transformers,which captures long-range contextual information,has demonstrated significant potential in image segmentation.However,their ability to learn local,contextual relationships between pixels requires further improvement.Previous methods face challenges in efficiently managing multi-scale fea-tures of different granularities from the encoder backbone,leaving room for improvement in their global representation and feature extraction capabilities.To address these challenges,we propose a novel Decoder with Multi-Head Feature Receptors(DMHFR),which receives multi-scale features from the encoder backbone and organizes them into three feature groups with different granularities:coarse,fine-grained,and full set.These groups are subsequently processed by Multi-Head Feature Receptors(MHFRs)after feature capture and modeling operations.MHFRs include two Three-Head Feature Receptors(THFRs)and one Four-Head Feature Receptor(FHFR).Each group of features is passed through these MHFRs and then fed into axial transformers,which help the model capture long-range dependencies within the features.The three MHFRs produce three distinct feature outputs.The output from the FHFR serves as auxiliary auxiliary features in the prediction head,and the prediction output and their losses will eventually be aggregated.Experimental results show that the Transformer using DMHFR outperforms 15 state of the arts(SOTA)methods on five public datasets.Specifically,it achieved significant improvements in mean DICE scores over the classic Parallel Reverse Attention Network(PraNet)method,with gains of 4.1%,2.2%,1.4%,8.9%,and 16.3%on the CVC-ClinicDB,Kvasir-SEG,CVC-T,CVC-ColonDB,and ETIS-LaribPolypDB datasets,respectively.
基金supported by Continuous-Support Basic Scientific Research Project(No.BJ010261223282)the National Natural Science Foundation of China(Nos.U2167201,11975318)+2 种基金the State Key Laboratory of Nuclear Physics and TechnologyPeking University(No.NPT2023KFY01)the Research and Development Project of China National Nuclear Corporation。
文摘The prompt fission neutron spectrum(PFNS)is a key nuclear data quantity that is of particular interest and plays a crucial role in understanding and modeling fission processes.An array comprising 48 liquid scintillation detectors and a parallelplate avalanche counter(PPAC)was developed at the China Institute of Atomic Energy(CIAE)to measure the PFNS of actinide nuclei.Efficiency and energy calibrations were performed for all the liquid scintillators,and their efficiencies were consistently found to be better than 5%.The time resolutions of the PPAC and liquid scintillators were measured to be 1.08 ns and 1.16 ns using~(252)Cf and~(207)Bi sources,respectively.The pulse shape discrimination of the liquid scintillator was utilized to identify neutron andγsignals on an event-by-event basis,and the figure of merit was deduced as 1.12 at a 200 ke Vee threshold.The contribution to the PFNS from multiple scattered neutrons was evaluated via Geant4 simulations,and those originating from the environment were found to be comparable to the crosstalk between the detectors.The neutron efficiency of the entire detection array was calibrated using a~(252)Cf spontaneous fission source and was demonstrated to be consistent with the Geant4 simulation results,which verified the reliability of the detection array.
