Scene graph prediction has emerged as a critical task in computer vision,focusing on transforming complex visual scenes into structured representations by identifying objects,their attributes,and the relationships amo...Scene graph prediction has emerged as a critical task in computer vision,focusing on transforming complex visual scenes into structured representations by identifying objects,their attributes,and the relationships among them.Extending this to 3D semantic scene graph(3DSSG)prediction introduces an additional layer of complexity because it requires the processing of point-cloud data to accurately capture the spatial and volumetric characteristics of a scene.A significant challenge in 3DSSG is the long-tailed distribution of object and relationship labels,causing certain classes to be severely underrepresented and suboptimal performance in these rare categories.To address this,we proposed a fusion prototypical network(FPN),which combines the strengths of conventional neural networks for 3DSSG with a Prototypical Network.The former are known for their ability to handle complex scene graph predictions while the latter excels in few-shot learning scenarios.By leveraging this fusion,our approach enhances the overall prediction accuracy and substantially improves the handling of underrepresented labels.Through extensive experiments using the 3DSSG dataset,we demonstrated that the FPN achieves state-of-the-art performance in 3D scene graph prediction as a single model and effectively mitigates the impact of the long-tailed distribution,providing a more balanced and comprehensive understanding of complex 3D environments.展开更多
In order to improve the models capability in expressing features during few-shot learning,a multi-scale features prototypical network(MS-PN)algorithm is proposed.The metric learning algo-rithm is employed to extract i...In order to improve the models capability in expressing features during few-shot learning,a multi-scale features prototypical network(MS-PN)algorithm is proposed.The metric learning algo-rithm is employed to extract image features and project them into a feature space,thus evaluating the similarity between samples based on their relative distances within the metric space.To sufficiently extract feature information from limited sample data and mitigate the impact of constrained data vol-ume,a multi-scale feature extraction network is presented to capture data features at various scales during the process of image feature extraction.Additionally,the position of the prototype is fine-tuned by assigning weights to data points to mitigate the influence of outliers on the experiment.The loss function integrates contrastive loss and label-smoothing to bring similar data points closer and separate dissimilar data points within the metric space.Experimental evaluations are conducted on small-sample datasets mini-ImageNet and CUB200-2011.The method in this paper can achieve higher classification accuracy.Specifically,in the 5-way 1-shot experiment,classification accuracy reaches 50.13%and 66.79%respectively on these two datasets.Moreover,in the 5-way 5-shot ex-periment,accuracy of 66.79%and 85.91%are observed,respectively.展开更多
Existing almost deep learning methods rely on a large amount of annotated data, so they are inappropriate for forest fire smoke detection with limited data. In this paper, a novel hybrid attention-based few-shot learn...Existing almost deep learning methods rely on a large amount of annotated data, so they are inappropriate for forest fire smoke detection with limited data. In this paper, a novel hybrid attention-based few-shot learning method, named Attention-Based Prototypical Network, is proposed for forest fire smoke detection. Specifically, feature extraction network, which consists of convolutional block attention module, could extract high-level and discriminative features and further decrease the false alarm rate resulting from suspected smoke areas. Moreover, we design a metalearning module to alleviate the overfitting issue caused by limited smoke images, and the meta-learning network enables achieving effective detection via comparing the distance between the class prototype of support images and the features of query images. A series of experiments on forest fire smoke datasets and miniImageNet dataset testify that the proposed method is superior to state-of-the-art few-shot learning approaches.展开更多
Few-shot Learning algorithms can be effectively applied to fields where certain categories have only a small amount of data or a small amount of labeled data,such as medical images,terrorist surveillance,and so on.The...Few-shot Learning algorithms can be effectively applied to fields where certain categories have only a small amount of data or a small amount of labeled data,such as medical images,terrorist surveillance,and so on.The Metric Learning in the Few-shot Learning algorithmis classified by measuring the similarity between the classified samples and the unclassified samples.This paper improves the Prototypical Network in the Metric Learning,and changes its core metric function to Manhattan distance.The Convolutional Neural Network of the embedded module is changed,and mechanisms such as average pooling and Dropout are added.Through comparative experiments,it is found that thismodel can converge in a small number of iterations(below 15,000 episodes),and its performance exceeds algorithms such asMAML.Research shows that replacingManhattan distance with Euclidean distance can effectively improve the classification effect of the Prototypical Network,and mechanisms such as average pooling and Dropout can also effectively improve the model.展开更多
This paper introduces a new approach dedicated to the Ontology Personalization. Inspired by works in Cognitive Psychology, our work is based on a process which aims at capturing the user-sensitive relevance of the cat...This paper introduces a new approach dedicated to the Ontology Personalization. Inspired by works in Cognitive Psychology, our work is based on a process which aims at capturing the user-sensitive relevance of the categorization process, that is the one which is really perceived by the end-user. Practically, this process consists in decorating the Specialization/Generalization links (i.e. the is-a links) of the hierarchy of concepts with 2 gradients. The goal of the first gradient, called Conceptual Prototypicality Gradient, is to capture the user-sensitive relevance of the categorization process, that is the one which is perceived by the end-user. As this gradient is defined according to the three aspects of the semiotic triangle (i.e. intentional, extensional and expressional dimension), we call it Semiotic based Prototypicality Gradient. The objective of the second gradient, called Lexical Prototypicality Gradient, is to capture the user-sensitive relevance of the lexicalization process, i.e. the definition of a set of terms used to denote a concept. These gradients enrich the initial formal semantics of an ontology by adding a pragmatics defined according to a context of use which depends on parameters like culture, educational background and/or emotional context of the end-user. This paper also introduces a new similarity measure also defined in the context of a semiotic-based approach. The first originality of this measure, called SEMIOSEM, is to consider the three semiotic dimensions of the conceptualization underlying an ontology. Thus, SEMIOSEM aims at aggregating and improving existing extensional-based and intentional-based measures. The second originality of this measure is to be context-sensitive, and in particular user-sensitive. This makes SEMIOSEM more flexible, more robust and more close to the end-user’s judgment than the other similarity measures which are usually only based on one aspect of a conceptualization and never take the end-user’s perceptions and purposes into account.展开更多
Starting from the traditional analysis of English intonation, the article discusses the Halliday＇s, Jackendoff＇s, and Brazil＇s views upon intonation. Then it explores the multiple meanings and its metonymic pattern...Starting from the traditional analysis of English intonation, the article discusses the Halliday＇s, Jackendoff＇s, and Brazil＇s views upon intonation. Then it explores the multiple meanings and its metonymic patterns of English high key in terms of the prototype theory.展开更多
This paper applies prototype theory to explain the motivation of polysemy. There are mainly 3 types of meaning model of polysemy, namely, radiation, concatenation and integrated model. According to prototype theory, i...This paper applies prototype theory to explain the motivation of polysemy. There are mainly 3 types of meaning model of polysemy, namely, radiation, concatenation and integrated model. According to prototype theory, in the semantic category formed by a polysemic word, category members are determined by prototype. They are the result of development from prototype to boundary. Connected by a network of overlapping similarities （i.e., family resemblances）, category members present different degrees of prototypicality, but not all of them can represent the category. Only the prototype can fully embody the category. With the extension of the semantic category, the boundary of the category is fuzzy and begins to intersect another semantic category.展开更多
Predicting future heart rate(HR)not only helps in detecting abnormal heart rhythms but also provides timely support for downstream health monitoring services.Existing methods for HR prediction encounter challenges,esp...Predicting future heart rate(HR)not only helps in detecting abnormal heart rhythms but also provides timely support for downstream health monitoring services.Existing methods for HR prediction encounter challenges,especially concerning privacy protection and data heterogeneity.To address these challenges,this paper proposes a novel HR prediction framework,PCFedH,which leverages personalized federated learning and prototypical contrastive learning to achieve stable clustering results and more accurate predictions.PCFedH contains two core modules:a prototypical contrastive learning-based federated clustering module,which characterizes data heterogeneity and enhances HR representation to facilitate more effective clustering,and a two-phase soft clustered federated learning module,which enables personalized performance improvements for each local model based on stable clustering results.Experimental results on two real-world datasets demonstrate the superiority of our approach over state-of-the-art methods,achieving an average reduction of 3.1%in the mean squared error across both datasets.Additionally,we conduct comprehensive experiments to empirically validate the effectiveness of the key components in the proposed method.Among these,the personalization component is identified as the most crucial aspect of our design,indicating its substantial impact on overall performance.展开更多
Partial label learning(PLL)learns under label ambiguity where each training instance is annotated with a set of candidate labels,among which only one is the ground-truth label.Recent advances showed that PLL can be pr...Partial label learning(PLL)learns under label ambiguity where each training instance is annotated with a set of candidate labels,among which only one is the ground-truth label.Recent advances showed that PLL can be promoted by combining label disambiguation with representation learning coherently,which achieved state-of-the-art performance.However,most of the existing deep PLL methods over-emphasize pulling the inaccurate pseudo-label-induced positive samples and fail to achieve a balance between the intra-class compactness and the inter-class separability,thus leading to a sub-optimal representation space.In this paper,we solve this issue by taking into account the pure negative supervision information which can be extracted perfectly from the non-candidate label set.Methodologically,we propose a novel framework Negative Prototypical Contrasting(NPC).The optimization objective of NPC contrasts each instance with its candidate prototypes against its negative prototypes,aiming at a sufficiently distinguishable representation space.Based on the learned representations,the label disambiguation process is performed in a moving-average style.Theoretically,we show that the objective of NPC is equivalent to solving a constrained maximum likelihood optimization.We also justify applying the moving average from the stochastic expectation-maximization perspective.Empirically,extensive experiments demonstrate that the proposed NPC method achieves state-of-the-art classification performance on various datasets,and even competes with its supervised counterparts.展开更多
Air quality estimation assesses the pollution level in the air,supports public health warnings,and is a valuable tool in environmental management.Although air sensors have proven helpful in this task,sensors are often...Air quality estimation assesses the pollution level in the air,supports public health warnings,and is a valuable tool in environmental management.Although air sensors have proven helpful in this task,sensors are often expensive and difficult to install,while cameras are becoming more popular and accessible,from which images can be collected as data for deep learning models to solve the above task.This leads to another problem:several labeled images are needed to achieve high accuracy when deep-learningmodels predict air quality.In this research,we have threemain contributions:(1)Collect and publish an air quality estimation dataset,namely PTIT_AQED,including environmental image data and air quality;(2)Propose a deep learning model to predict air quality with few data,called PTIT_FAQE(PTIT Few-shot air quality estimation).We build PTIT_FAQE based on EfficientNet-a CNN architecture that ensures high performance in deep learning applications and Few-shot Learning with Prototypical Networks.This helps the model use only a fewtraining data but still achieve high accuracy in air quality estimation.And(3)conduct experiments to prove the superiority of PTIT_FAQE compared to other studies on both PTIT_AQED and APIN datasets.The results show that our model achieves an accuracy of 0.9278 and an F1-Score of 0.9139 on the PTIT_AQED dataset and an accuracy of 0.9467 and an F1-Score of 0.9371 on the APIN dataset,which demonstrate a significant performance improvement compared to previous studies.We also conduct detailed experiments to evaluate the impact of each component on model performance.展开更多
无监督异常检测因只需要正常样本进行训练而被广泛应用于工业质检等领域。直接将现有的单类别异常检测方法应用到多类别异常检测中会导致性能显著下降,其中基于知识蒸馏的异常检测方法将预训练的教师模型关于正常样本的特征知识蒸馏到...无监督异常检测因只需要正常样本进行训练而被广泛应用于工业质检等领域。直接将现有的单类别异常检测方法应用到多类别异常检测中会导致性能显著下降,其中基于知识蒸馏的异常检测方法将预训练的教师模型关于正常样本的特征知识蒸馏到学生模型中,然而它们在多类别异常检测中存在无法保证学生模型只学习到正常样本知识的问题。文中提出一种基于反向知识蒸馏框架的无监督多类别异常检测方法(Prototype based Reverse Distillation,PRD),其通过Multi-class Normal Prototype模块和Sparse Prototype Recall训练策略来学习教师模型关于多类别正常样本特征的Prototype,并以此来过滤学生模型的输入特征,从而确保学生模型只学习到教师模型关于正常样本的特征知识。PRD在多种工业异常检测数据集上性能均超越了现有的SOTA方法,定性、定量和消融实验验证了PRD整体框架和内部模块的有效性。展开更多
This paper is based on the assumption that semantic categories vary from one language to another and prototypes of semantic categories are culture-specific. Prototypicality has its effects in acquisition of second for...This paper is based on the assumption that semantic categories vary from one language to another and prototypes of semantic categories are culture-specific. Prototypicality has its effects in acquisition of second foreign language vocabulary. The author hypothesizes that focal member words and basic level words are acquired earlier than non-focal member words and words on other levels of lexical hierarchies when prototypes of the target language correspond to those of the mother tongue in the same semantic categories. The results of a vocabulary test on beginner, intermediate and advanced learners of English as foreign language confirm the hypothesis. The paper concludes that learning a second foreign language means, to an extent, learning to categorize the worm in a different way, and prototypic effects should be given due attention in vocabulary instruction.展开更多
Radio frequency(RF)cavities for advanced storage rings,also known as diffraction-limited storage rings,are under development.To this end,a competitive and promising approach involves normal-conducting continuous wave ...Radio frequency(RF)cavities for advanced storage rings,also known as diffraction-limited storage rings,are under development.To this end,a competitive and promising approach involves normal-conducting continuous wave technology.The design and preliminary test of a 499.654 MHz RF cavity for the Wuhan Advanced Light Source(WALS)based on specific beam parameters were conducted at the SSRF.Multi-objective evolutionary algorithms have been utilized to optimize RF properties,such as the power loss and power density,resulting in better performance in the continuous wave mode.Further improvements were made to suppress multipacting effects in the working area.To operate stably with the beam,higher-order mode dampers were applied to better address the coupling bunch instability than in previous designs,along with thermal analysis to achieve the desired RF performance.Comprehensive simulation studies demonstrated the stable operation of the RF cavity at the defined beam parameters in the WALS design.A prototype RF cavity was then developed,and the RF performance results in a low-power test showed good agreement with the design and simulation,exhibiting readiness for high-power experiments and operation.展开更多
There are many traditional villages with well-preserved architectural types and images in the Jingmai Mountain,Yunnan Province.Through field investigations in traditional villages in the research area,this study appli...There are many traditional villages with well-preserved architectural types and images in the Jingmai Mountain,Yunnan Province.Through field investigations in traditional villages in the research area,this study applied the architectural typology,analyzed Nuogang Village of the Dai Nationality and Wengji Village of the Bulang Nationality from 3 perspectives of“point,line and surface”,explored the characteristics of village,architecture and landscape,and extracted the“prototypes”,tried to figure out the problems of the villages and then propose corresponding protection strategies,so as to support the preservation,renovation,improvement and utilization of traditional villages.展开更多
Aircraft have received much attention because of their capability to adapt to various flight environments and complex missions.The nose cone is one of the key elements in optimising the aerodynamic shape of aircraft.A...Aircraft have received much attention because of their capability to adapt to various flight environments and complex missions.The nose cone is one of the key elements in optimising the aerodynamic shape of aircraft.A morphing nose cone(MNC)driven by a biomimetic 4-3R1U&3R sparallel mechanism is proposed in this study.Based on screw theory,the parallel mechanism’s configuration is determined,and the structure’s full-cycle degrees of freedom are concurrently confirmed.Examples in the paper demonstrate the viability of the structure by configuration synthesis,and diagrams also show the chains.This MNC is modelled after the structural design of the cicada’s abdomen and can be extended,contracted and bent.It can actively adjust its shape in response to change in the flight environments,thereby aerodynamic performance and enhancing the aircraft’s multi-mission capabilities.A scaled-down prototype is created to verify the deformation capacity of the MNC meeting the engineering requirements.Results show that the extension ratio is 36.7%,and the bending angle is 21.7°,which is better than expected.The relative error value is within a reasonable range and the extension process is incredibly stable.This research proposes new perspectives for the design of MNCs.展开更多
Since the introduction of vision Transformers into the computer vision field,many vision tasks such as semantic segmentation tasks,have undergone radical changes.Although Transformer enhances the correlation of each l...Since the introduction of vision Transformers into the computer vision field,many vision tasks such as semantic segmentation tasks,have undergone radical changes.Although Transformer enhances the correlation of each local feature of an image object in the hidden space through the attention mechanism,it is difficult for a segmentation head to accomplish the mask prediction for dense embedding of multi-category and multi-local features.We present patch prototype vision Transformer(PPFormer),a Transformer architecture for semantic segmentation based on knowledge-embedded patch prototypes.1)The hierarchical Transformer encoder can generate multi-scale and multi-layered patch features including seamless patch projection to obtain information of multiscale patches,and feature-clustered self-attention to enhance the interplay of multi-layered visual information with implicit position encodes.2)PPFormer utilizes a non-parametric prototype decoder to extract region observations which represent significant parts of the objects by unlearnable patch prototypes and then calculate similarity between patch prototypes and pixel embeddings.The proposed contrasting patch prototype alignment module,which uses new patch prototypes to update prototype bank,effectively maintains class boundaries for prototypes.For different application scenarios,we have launched PPFormer-S,PPFormer-M and PPFormer-L by expanding the scale.Experimental results demonstrate that PPFormer can outperform fully convolutional networks(FCN)-and attention-based semantic segmentation models on the PASCAL VOC 2012,ADE20k,and Cityscapes datasets.展开更多
A team of researchers from the University of Science and Technology of China(USTC)of the Chinese Academy of Sciences(CAS)and its partners have made significant advancements in random quantum circuit sampling with Zuch...A team of researchers from the University of Science and Technology of China(USTC)of the Chinese Academy of Sciences(CAS)and its partners have made significant advancements in random quantum circuit sampling with Zuchongzhi-3,a superconducting quantum computing prototype featuring 105 qubits and 182 couplers.展开更多
To enable simultaneous transmit and receive(STAR)on the same frequency in a densely deployed space with multi-interference sources,this work proposes a digitally-assisted analog selfinterference cancellation method,wh...To enable simultaneous transmit and receive(STAR)on the same frequency in a densely deployed space with multi-interference sources,this work proposes a digitally-assisted analog selfinterference cancellation method,which can acquire reference signals through flexible wired/wireless switching access.Based on this method,the Minimum Mean Square Error algorithm with known channel state information is derived in detail,determining the upper limit of the cancellation performance,and the Adaptive Dithered Linear Search algorithm for real-time engineering cancellation is given.The correctness of theoretical analysis is verified by the practical self-interference channel measured by a vector network analyzer.Furthermore,we have designed and implemented the corresponding multiinterference cancellation prototype with the digitallyassisted structure,capable of handling multiple interferences(up to three)and supporting a large receive bandwidth of 100 MHz as well as a wide frequency coverage from 30 MHz to 3000 MHz.Prototype test results demonstrate that in the presence of three interferences,when the single interference bandwidth is 0.2/2/20 MHz(corresponding to the receive bandwidth of 2/20/100 MHz),the cancellation performance can reach 46/32/22 dB or more.展开更多
基金supported by the Glocal University 30 Project Fund of Gyeongsang National University in 2025.
文摘Scene graph prediction has emerged as a critical task in computer vision,focusing on transforming complex visual scenes into structured representations by identifying objects,their attributes,and the relationships among them.Extending this to 3D semantic scene graph(3DSSG)prediction introduces an additional layer of complexity because it requires the processing of point-cloud data to accurately capture the spatial and volumetric characteristics of a scene.A significant challenge in 3DSSG is the long-tailed distribution of object and relationship labels,causing certain classes to be severely underrepresented and suboptimal performance in these rare categories.To address this,we proposed a fusion prototypical network(FPN),which combines the strengths of conventional neural networks for 3DSSG with a Prototypical Network.The former are known for their ability to handle complex scene graph predictions while the latter excels in few-shot learning scenarios.By leveraging this fusion,our approach enhances the overall prediction accuracy and substantially improves the handling of underrepresented labels.Through extensive experiments using the 3DSSG dataset,we demonstrated that the FPN achieves state-of-the-art performance in 3D scene graph prediction as a single model and effectively mitigates the impact of the long-tailed distribution,providing a more balanced and comprehensive understanding of complex 3D environments.
基金the Scientific Research Foundation of Liaoning Provincial Department of Education(No.LJKZ0139)the Program for Liaoning Excellent Talents in University(No.LR15045).
文摘In order to improve the models capability in expressing features during few-shot learning,a multi-scale features prototypical network(MS-PN)algorithm is proposed.The metric learning algo-rithm is employed to extract image features and project them into a feature space,thus evaluating the similarity between samples based on their relative distances within the metric space.To sufficiently extract feature information from limited sample data and mitigate the impact of constrained data vol-ume,a multi-scale feature extraction network is presented to capture data features at various scales during the process of image feature extraction.Additionally,the position of the prototype is fine-tuned by assigning weights to data points to mitigate the influence of outliers on the experiment.The loss function integrates contrastive loss and label-smoothing to bring similar data points closer and separate dissimilar data points within the metric space.Experimental evaluations are conducted on small-sample datasets mini-ImageNet and CUB200-2011.The method in this paper can achieve higher classification accuracy.Specifically,in the 5-way 1-shot experiment,classification accuracy reaches 50.13%and 66.79%respectively on these two datasets.Moreover,in the 5-way 5-shot ex-periment,accuracy of 66.79%and 85.91%are observed,respectively.
基金The work was supported by the National Key R&D Program of China(Grant No.2020YFC1511601)Fundamental Research Funds for the Central Universities(Grant No.2019SHFWLC01).
文摘Existing almost deep learning methods rely on a large amount of annotated data, so they are inappropriate for forest fire smoke detection with limited data. In this paper, a novel hybrid attention-based few-shot learning method, named Attention-Based Prototypical Network, is proposed for forest fire smoke detection. Specifically, feature extraction network, which consists of convolutional block attention module, could extract high-level and discriminative features and further decrease the false alarm rate resulting from suspected smoke areas. Moreover, we design a metalearning module to alleviate the overfitting issue caused by limited smoke images, and the meta-learning network enables achieving effective detection via comparing the distance between the class prototype of support images and the features of query images. A series of experiments on forest fire smoke datasets and miniImageNet dataset testify that the proposed method is superior to state-of-the-art few-shot learning approaches.
文摘Few-shot Learning algorithms can be effectively applied to fields where certain categories have only a small amount of data or a small amount of labeled data,such as medical images,terrorist surveillance,and so on.The Metric Learning in the Few-shot Learning algorithmis classified by measuring the similarity between the classified samples and the unclassified samples.This paper improves the Prototypical Network in the Metric Learning,and changes its core metric function to Manhattan distance.The Convolutional Neural Network of the embedded module is changed,and mechanisms such as average pooling and Dropout are added.Through comparative experiments,it is found that thismodel can converge in a small number of iterations(below 15,000 episodes),and its performance exceeds algorithms such asMAML.Research shows that replacingManhattan distance with Euclidean distance can effectively improve the classification effect of the Prototypical Network,and mechanisms such as average pooling and Dropout can also effectively improve the model.
文摘This paper introduces a new approach dedicated to the Ontology Personalization. Inspired by works in Cognitive Psychology, our work is based on a process which aims at capturing the user-sensitive relevance of the categorization process, that is the one which is really perceived by the end-user. Practically, this process consists in decorating the Specialization/Generalization links (i.e. the is-a links) of the hierarchy of concepts with 2 gradients. The goal of the first gradient, called Conceptual Prototypicality Gradient, is to capture the user-sensitive relevance of the categorization process, that is the one which is perceived by the end-user. As this gradient is defined according to the three aspects of the semiotic triangle (i.e. intentional, extensional and expressional dimension), we call it Semiotic based Prototypicality Gradient. The objective of the second gradient, called Lexical Prototypicality Gradient, is to capture the user-sensitive relevance of the lexicalization process, i.e. the definition of a set of terms used to denote a concept. These gradients enrich the initial formal semantics of an ontology by adding a pragmatics defined according to a context of use which depends on parameters like culture, educational background and/or emotional context of the end-user. This paper also introduces a new similarity measure also defined in the context of a semiotic-based approach. The first originality of this measure, called SEMIOSEM, is to consider the three semiotic dimensions of the conceptualization underlying an ontology. Thus, SEMIOSEM aims at aggregating and improving existing extensional-based and intentional-based measures. The second originality of this measure is to be context-sensitive, and in particular user-sensitive. This makes SEMIOSEM more flexible, more robust and more close to the end-user’s judgment than the other similarity measures which are usually only based on one aspect of a conceptualization and never take the end-user’s perceptions and purposes into account.
文摘Starting from the traditional analysis of English intonation, the article discusses the Halliday＇s, Jackendoff＇s, and Brazil＇s views upon intonation. Then it explores the multiple meanings and its metonymic patterns of English high key in terms of the prototype theory.
文摘This paper applies prototype theory to explain the motivation of polysemy. There are mainly 3 types of meaning model of polysemy, namely, radiation, concatenation and integrated model. According to prototype theory, in the semantic category formed by a polysemic word, category members are determined by prototype. They are the result of development from prototype to boundary. Connected by a network of overlapping similarities （i.e., family resemblances）, category members present different degrees of prototypicality, but not all of them can represent the category. Only the prototype can fully embody the category. With the extension of the semantic category, the boundary of the category is fuzzy and begins to intersect another semantic category.
基金supported by the National Natural Science Foundation of China(Nos.62102094 and 62072115)the Shanghai Science and Technology Innovation Action Plan Project(No.22510713600)the NIO University Programme,and the Nordic University Cooperation on Edge Intelligence(No.168043)。
文摘Predicting future heart rate(HR)not only helps in detecting abnormal heart rhythms but also provides timely support for downstream health monitoring services.Existing methods for HR prediction encounter challenges,especially concerning privacy protection and data heterogeneity.To address these challenges,this paper proposes a novel HR prediction framework,PCFedH,which leverages personalized federated learning and prototypical contrastive learning to achieve stable clustering results and more accurate predictions.PCFedH contains two core modules:a prototypical contrastive learning-based federated clustering module,which characterizes data heterogeneity and enhances HR representation to facilitate more effective clustering,and a two-phase soft clustered federated learning module,which enables personalized performance improvements for each local model based on stable clustering results.Experimental results on two real-world datasets demonstrate the superiority of our approach over state-of-the-art methods,achieving an average reduction of 3.1%in the mean squared error across both datasets.Additionally,we conduct comprehensive experiments to empirically validate the effectiveness of the key components in the proposed method.Among these,the personalization component is identified as the most crucial aspect of our design,indicating its substantial impact on overall performance.
基金supported by the National Natural Science Foundation of China under Grant No.U23A20468.
文摘Partial label learning(PLL)learns under label ambiguity where each training instance is annotated with a set of candidate labels,among which only one is the ground-truth label.Recent advances showed that PLL can be promoted by combining label disambiguation with representation learning coherently,which achieved state-of-the-art performance.However,most of the existing deep PLL methods over-emphasize pulling the inaccurate pseudo-label-induced positive samples and fail to achieve a balance between the intra-class compactness and the inter-class separability,thus leading to a sub-optimal representation space.In this paper,we solve this issue by taking into account the pure negative supervision information which can be extracted perfectly from the non-candidate label set.Methodologically,we propose a novel framework Negative Prototypical Contrasting(NPC).The optimization objective of NPC contrasts each instance with its candidate prototypes against its negative prototypes,aiming at a sufficiently distinguishable representation space.Based on the learned representations,the label disambiguation process is performed in a moving-average style.Theoretically,we show that the objective of NPC is equivalent to solving a constrained maximum likelihood optimization.We also justify applying the moving average from the stochastic expectation-maximization perspective.Empirically,extensive experiments demonstrate that the proposed NPC method achieves state-of-the-art classification performance on various datasets,and even competes with its supervised counterparts.
文摘Air quality estimation assesses the pollution level in the air,supports public health warnings,and is a valuable tool in environmental management.Although air sensors have proven helpful in this task,sensors are often expensive and difficult to install,while cameras are becoming more popular and accessible,from which images can be collected as data for deep learning models to solve the above task.This leads to another problem:several labeled images are needed to achieve high accuracy when deep-learningmodels predict air quality.In this research,we have threemain contributions:(1)Collect and publish an air quality estimation dataset,namely PTIT_AQED,including environmental image data and air quality;(2)Propose a deep learning model to predict air quality with few data,called PTIT_FAQE(PTIT Few-shot air quality estimation).We build PTIT_FAQE based on EfficientNet-a CNN architecture that ensures high performance in deep learning applications and Few-shot Learning with Prototypical Networks.This helps the model use only a fewtraining data but still achieve high accuracy in air quality estimation.And(3)conduct experiments to prove the superiority of PTIT_FAQE compared to other studies on both PTIT_AQED and APIN datasets.The results show that our model achieves an accuracy of 0.9278 and an F1-Score of 0.9139 on the PTIT_AQED dataset and an accuracy of 0.9467 and an F1-Score of 0.9371 on the APIN dataset,which demonstrate a significant performance improvement compared to previous studies.We also conduct detailed experiments to evaluate the impact of each component on model performance.
文摘无监督异常检测因只需要正常样本进行训练而被广泛应用于工业质检等领域。直接将现有的单类别异常检测方法应用到多类别异常检测中会导致性能显著下降,其中基于知识蒸馏的异常检测方法将预训练的教师模型关于正常样本的特征知识蒸馏到学生模型中,然而它们在多类别异常检测中存在无法保证学生模型只学习到正常样本知识的问题。文中提出一种基于反向知识蒸馏框架的无监督多类别异常检测方法(Prototype based Reverse Distillation,PRD),其通过Multi-class Normal Prototype模块和Sparse Prototype Recall训练策略来学习教师模型关于多类别正常样本特征的Prototype,并以此来过滤学生模型的输入特征,从而确保学生模型只学习到教师模型关于正常样本的特征知识。PRD在多种工业异常检测数据集上性能均超越了现有的SOTA方法,定性、定量和消融实验验证了PRD整体框架和内部模块的有效性。
文摘This paper is based on the assumption that semantic categories vary from one language to another and prototypes of semantic categories are culture-specific. Prototypicality has its effects in acquisition of second foreign language vocabulary. The author hypothesizes that focal member words and basic level words are acquired earlier than non-focal member words and words on other levels of lexical hierarchies when prototypes of the target language correspond to those of the mother tongue in the same semantic categories. The results of a vocabulary test on beginner, intermediate and advanced learners of English as foreign language confirm the hypothesis. The paper concludes that learning a second foreign language means, to an extent, learning to categorize the worm in a different way, and prototypic effects should be given due attention in vocabulary instruction.
基金supported by National Natural Science Foundation of China(Nos.12222513,12105345,12175292,and No.12405178)。
文摘Radio frequency(RF)cavities for advanced storage rings,also known as diffraction-limited storage rings,are under development.To this end,a competitive and promising approach involves normal-conducting continuous wave technology.The design and preliminary test of a 499.654 MHz RF cavity for the Wuhan Advanced Light Source(WALS)based on specific beam parameters were conducted at the SSRF.Multi-objective evolutionary algorithms have been utilized to optimize RF properties,such as the power loss and power density,resulting in better performance in the continuous wave mode.Further improvements were made to suppress multipacting effects in the working area.To operate stably with the beam,higher-order mode dampers were applied to better address the coupling bunch instability than in previous designs,along with thermal analysis to achieve the desired RF performance.Comprehensive simulation studies demonstrated the stable operation of the RF cavity at the defined beam parameters in the WALS design.A prototype RF cavity was then developed,and the RF performance results in a low-power test showed good agreement with the design and simulation,exhibiting readiness for high-power experiments and operation.
文摘There are many traditional villages with well-preserved architectural types and images in the Jingmai Mountain,Yunnan Province.Through field investigations in traditional villages in the research area,this study applied the architectural typology,analyzed Nuogang Village of the Dai Nationality and Wengji Village of the Bulang Nationality from 3 perspectives of“point,line and surface”,explored the characteristics of village,architecture and landscape,and extracted the“prototypes”,tried to figure out the problems of the villages and then propose corresponding protection strategies,so as to support the preservation,renovation,improvement and utilization of traditional villages.
基金Supported by Hebei Provincial Natural Science Foundation(Grant Nos.E2024203052,E2024203105)National Natural Science Foundation of China(Grant No.52375028)Science and Technology Project of Hebei Education Department(Grant No.QN2023206).
文摘Aircraft have received much attention because of their capability to adapt to various flight environments and complex missions.The nose cone is one of the key elements in optimising the aerodynamic shape of aircraft.A morphing nose cone(MNC)driven by a biomimetic 4-3R1U&3R sparallel mechanism is proposed in this study.Based on screw theory,the parallel mechanism’s configuration is determined,and the structure’s full-cycle degrees of freedom are concurrently confirmed.Examples in the paper demonstrate the viability of the structure by configuration synthesis,and diagrams also show the chains.This MNC is modelled after the structural design of the cicada’s abdomen and can be extended,contracted and bent.It can actively adjust its shape in response to change in the flight environments,thereby aerodynamic performance and enhancing the aircraft’s multi-mission capabilities.A scaled-down prototype is created to verify the deformation capacity of the MNC meeting the engineering requirements.Results show that the extension ratio is 36.7%,and the bending angle is 21.7°,which is better than expected.The relative error value is within a reasonable range and the extension process is incredibly stable.This research proposes new perspectives for the design of MNCs.
基金supported in part by the Gansu Haizhi Characteristic Demonstration Project(No.GSHZTS2022-2).
文摘Since the introduction of vision Transformers into the computer vision field,many vision tasks such as semantic segmentation tasks,have undergone radical changes.Although Transformer enhances the correlation of each local feature of an image object in the hidden space through the attention mechanism,it is difficult for a segmentation head to accomplish the mask prediction for dense embedding of multi-category and multi-local features.We present patch prototype vision Transformer(PPFormer),a Transformer architecture for semantic segmentation based on knowledge-embedded patch prototypes.1)The hierarchical Transformer encoder can generate multi-scale and multi-layered patch features including seamless patch projection to obtain information of multiscale patches,and feature-clustered self-attention to enhance the interplay of multi-layered visual information with implicit position encodes.2)PPFormer utilizes a non-parametric prototype decoder to extract region observations which represent significant parts of the objects by unlearnable patch prototypes and then calculate similarity between patch prototypes and pixel embeddings.The proposed contrasting patch prototype alignment module,which uses new patch prototypes to update prototype bank,effectively maintains class boundaries for prototypes.For different application scenarios,we have launched PPFormer-S,PPFormer-M and PPFormer-L by expanding the scale.Experimental results demonstrate that PPFormer can outperform fully convolutional networks(FCN)-and attention-based semantic segmentation models on the PASCAL VOC 2012,ADE20k,and Cityscapes datasets.
文摘A team of researchers from the University of Science and Technology of China(USTC)of the Chinese Academy of Sciences(CAS)and its partners have made significant advancements in random quantum circuit sampling with Zuchongzhi-3,a superconducting quantum computing prototype featuring 105 qubits and 182 couplers.
基金supported in part by the National Natural Science Foundation of China under Grant 62071094in part by the National Key Laboratory of Wireless Communications Foundation under Grant IFN202402in part by the Postdoctoral Fellowship Program(Grade C)of China Postdoctoral Science Foundation under Grant GZC20240217.
文摘To enable simultaneous transmit and receive(STAR)on the same frequency in a densely deployed space with multi-interference sources,this work proposes a digitally-assisted analog selfinterference cancellation method,which can acquire reference signals through flexible wired/wireless switching access.Based on this method,the Minimum Mean Square Error algorithm with known channel state information is derived in detail,determining the upper limit of the cancellation performance,and the Adaptive Dithered Linear Search algorithm for real-time engineering cancellation is given.The correctness of theoretical analysis is verified by the practical self-interference channel measured by a vector network analyzer.Furthermore,we have designed and implemented the corresponding multiinterference cancellation prototype with the digitallyassisted structure,capable of handling multiple interferences(up to three)and supporting a large receive bandwidth of 100 MHz as well as a wide frequency coverage from 30 MHz to 3000 MHz.Prototype test results demonstrate that in the presence of three interferences,when the single interference bandwidth is 0.2/2/20 MHz(corresponding to the receive bandwidth of 2/20/100 MHz),the cancellation performance can reach 46/32/22 dB or more.
