Timely and accurate diagnosis of skin diseases is crucial as conventional methods are time-consuming and prone to errors.Traditional trial-and-error approaches often aggregate multiple models without optimization by r...Timely and accurate diagnosis of skin diseases is crucial as conventional methods are time-consuming and prone to errors.Traditional trial-and-error approaches often aggregate multiple models without optimization by resulting in suboptimal performance.To address these challenges,we propose a novel Squid Game OptimizationDimension Reduction-based Ensemble(SGO-DRE)method for the precise diagnosis of skin diseases.Our approach begins by selecting pre-trained models named MobileNetV1,DenseNet201,and Xception for robust feature extraction.These models are enhanced with dimension reduction blocks to improve efficiency.To tackle the aggregation problem of various models,we leverage the Squid Game Optimization(SGO)algorithm,which iteratively searches for the optimal weightage set to assign the appropriate weightage to each individual model within the proposed weighted average aggregation ensemble approach.The proposed ensemble method effectively utilizes the strengths of each model.We evaluated the proposed method using an 8-class skin disease dataset,a 6-class MSLD dataset,and a 4-class MSID dataset,achieving accuracies of 98.71%,96.34%,and 93.46%,respectively.Additionally,we employed visual tools like Grad-CAM,ROC curves,and Precision-Recall curves to interpret the decision making of models and assess its performance.These evaluations ensure that the proposed method not only provides robust results but also enhances interpretability and reliability in clinical decision-making.展开更多
With the increasing enlargement of network scale and the rapid development of network techniques, large numbers of the network applications begin to appear. Packet capture plays an important role as one basic techniqu...With the increasing enlargement of network scale and the rapid development of network techniques, large numbers of the network applications begin to appear. Packet capture plays an important role as one basic technique used in each field of the network applications. In a high-speed network, the heavy traffic of network transmission challenges the packet capture techniques. This paper does an in-depth analysis on the traditional packet capture mechanisms in Linux, and then measures the performance bottleneck in the process of packet capture. The methods for improving the packet capture performance are presented and an optimized packet capture scheme is also designed and implemented. The test demonstrates that the new packet capture mechanism (Libpacket) can greatly improve the packet capture performance of the network application systems in a high-speed network.展开更多
This paper shows the harm of harmonic in power system,compares the measures of normal digital filter and wavelet MARto afford reference to the detection and elimination in power system harmonic control.
The traveling salesman problem( TSP) is a well-known combinatorial optimization problem as well as an NP-complete problem. A dynamic multi-swarm particle swarm optimization and ant colony optimization( DMPSO-ACO) was ...The traveling salesman problem( TSP) is a well-known combinatorial optimization problem as well as an NP-complete problem. A dynamic multi-swarm particle swarm optimization and ant colony optimization( DMPSO-ACO) was presented for TSP.The DMPSO-ACO combined the exploration capabilities of the dynamic multi-swarm particle swarm optimizer( DMPSO) and the stochastic exploitation of the ant colony optimization( ACO) for solving the traveling salesman problem. In the proposed hybrid algorithm,firstly,the dynamic swarms,rapidity of the PSO was used to obtain a series of sub-optimal solutions through certain iterative times for adjusting the initial allocation of pheromone in ACO. Secondly,the positive feedback and high accuracy of the ACO were employed to solving whole problem. Finally,to verify the effectiveness and efficiency of the proposed hybrid algorithm,various scale benchmark problems were tested to demonstrate the potential of the proposed DMPSO-ACO algorithm. The results show that DMPSO-ACO is better in the search precision,convergence property and has strong ability to escape from the local sub-optima when compared with several other peer algorithms.展开更多
The scarcity of bandwidth in the radio spectrum has become more vital since the demand for more and more wireless applications has increased. Most of the spectrum bands have been allocated although many studies have s...The scarcity of bandwidth in the radio spectrum has become more vital since the demand for more and more wireless applications has increased. Most of the spectrum bands have been allocated although many studies have shown that these bands are significantly underutilized most of the time. The problem of unavailability of spectrum and inefficiency in its utilization has been smartly addressed by the cognitive radio （CR） technology which is an opportunistic network that senses the environment, observes the network changes, and then uses knowledge gained from the prior interaction with the network to make intelligent decisions by dynamically adapting their transmission characteristics. In this paper, some of the decentralized adaptive medium access control （MAC） protocols for CR networks have been critically analyzed, and a novel adaptive MAC protocol for CR networks, decentralized non-global MAC （DNG-MAC）, has been proposed. The results show the DNG-MAC outperforms other CR-MAC protocols in terms of time and energy efficiency.展开更多
A new ontology-based question expansion （OBQE） method is proposed for question similarity calculation in a frequently asked question （FAQ） answering system. Traditional question similarity calculation methods use ...A new ontology-based question expansion （OBQE） method is proposed for question similarity calculation in a frequently asked question （FAQ） answering system. Traditional question similarity calculation methods use ＂word＂ to compose question vector, that the semantic relations between words are ignored. OBQE takes the relation as an important part. The process of the new system is：① to build two-layered domain ontology referring to WordNet and domain corpse;② to expand question trunks into domain cases;③ to use domain case composed vector to calculate question similarity. The experimental result shows that the performance of question similarity calculation with OBQE is being improved.展开更多
A valid method of virtual scene depth calculating is put forward. In this method cameras rotate in three different viewpoints in the plane and we calculate the depth of panorama using three stitching cylinder panorama...A valid method of virtual scene depth calculating is put forward. In this method cameras rotate in three different viewpoints in the plane and we calculate the depth of panorama using three stitching cylinder panoramas. In the investigation, the column of panorama is regarded as a slot image. Using the conic intersected by the epipolar plane and the cylinder, we can obtain the pel-pendicularity disparity. In order to obtain dense correspondence fast and accurately, a new method of obtaining horizontal disparity using depth continuity is also put forward. It converts the problem of panorama dense correspondence to the problem of searching points in the conic. The occlusion problem is dealt with using three cylinders in the depth calculation. It is verified that this method is convenient, useful and efficient in calculating the depth of a virtual scene.展开更多
The mode of mobile computing originated from distributed computing and it has the un-idempotent operation property, therefore the deadlock detection algorithm designed for mobile computing systems will face challenges...The mode of mobile computing originated from distributed computing and it has the un-idempotent operation property, therefore the deadlock detection algorithm designed for mobile computing systems will face challenges with regard to correctness and high efficiency. This paper attempts a fundamental study of deadlock detection for the AND model of mobile computing systems. First, the existing deadlock detection algorithms for distributed systems are classified into the resource node dependent (RD) and the resource node independent (RI) categories, and their corresponding weaknesses are discussed. Afterwards a new RI algorithm based on the AND model of mobile computing system is presented. The novelties of our algorithm are that: 1) the blocked nodes inform their predecessors and successors simultaneously; 2) the detection messages (agents) hold the predecessors information of their originator; 3) no agent is stored midway. Additionally, the quit-inform scheme is introduced to treat the excessive victim quitting problem raised by the overlapped cycles. By these methods the proposed algorithm can detect a cycle of size n within n-2 steps and with (n^2-n-2)/2 agents. The performance of our algorithm is compared with the most competitive RD and RI algorithms for distributed systems on a mobile agent simulation platform. Experiment results point out that our algorithm outperforms the two algorithms under the vast majority of resource configurations and concurrent workloads. The correctness of the proposed algorithm is formally proven by the invariant verification technique.展开更多
The quantum theory application is a hot research area in recent years,especially the theory of quantum mechanics.In this paper,we focus on the research of image segmentation based on quantum mechanics.Firstly,the theo...The quantum theory application is a hot research area in recent years,especially the theory of quantum mechanics.In this paper,we focus on the research of image segmentation based on quantum mechanics.Firstly,the theory of quantum mechanics is introduced;afterwards,a review of image segmentation methods based on quantum mechanics is presented;and finally,the characteristics about the quantum mechanics applied to image processing are concluded.Two main research topics are discussed in this paper.One is to emphasize that quantum mechanics can be applied in different research areas,such as image segmentation,and the second is to conclude some methods in image segmentation and give some suggestions for possible novel methods by applying quantum mechanics theory.As a summary,this is a review paper which presents some methods based on the feasible theory in quantum mechanics aiming at achieving a better performance in image segmentation.展开更多
TikTok is one of the world’s most attractive and fastest-growing social media platforms.It has more than 2.6 billion downloads globally and over 100 million users in the US.The“secret weapon”is its unique methodolo...TikTok is one of the world’s most attractive and fastest-growing social media platforms.It has more than 2.6 billion downloads globally and over 100 million users in the US.The“secret weapon”is its unique methodology of discovering and delivering content.On the“For You”feed,TikTok mixes videos of both internet celebrities and newcomers,encourages high-quality creative content according to page views,and promotes new bloggers to share videos with users.The unique aspect is that anyone has the opportunity to spring into fame on the feed.Through TikTok’s recom-mendation algorithm,videos are continuously recommended to users with similar interests or attributes as video bloggers,thus allowing high-quality creative content to be disseminated quickly.The TikTok recommendation algorithm does not take video blogger’s fanbase or popularity into major consideration.In addition to the candidate video title,audio,and tags,the algorithm integrates the content of the user’s uploaded videos,and the categories of videos liked by the user.TikTok’s skills at enhancing user loyalty are impressive,which can not only accurately recommend videos of interest to users,but also assist them in expanding into new intersecting areas.展开更多
Relevance estimation is one of the core concerns of information retrieval(IR)studies.Although existing retrieval models gained much success in both deepening our understanding of information seeking behavior and build...Relevance estimation is one of the core concerns of information retrieval(IR)studies.Although existing retrieval models gained much success in both deepening our understanding of information seeking behavior and building effective retrieval systems,we have to admit that the models work in a rather different manner from how humans make relevance judgments.Users’information seeking behaviors involve complex cognitive processes,however,the majority of these behavior patterns are not considered in existing retrieval models.To bridge the gap between practical user behavior and retrieval model,it is essential to systematically investigate user cognitive behavior during relevance judgement and incorporate these heuristics into retrieval models.In this paper,we aim to formally define a set of basic user reading heuristics during relevance judgement and investigate their corresponding modeling strategies in retrieval models.Further experiments are conducted to evaluate the effectiveness of different reading heuristics for improving ranking performance.Based on a large-scale Web search dataset,we find that most reading heuristics can improve the performance of retrieval model and establish guidelines for improving the design of retrieval models with human-inspired heuristics.Our study sheds light on building retrieval model from the perspective of cognitive behavior.展开更多
Geometric graphs are a special kind of graph with geometric features,which are vital to model many scientific problems.Unlike generic graphs,geometric graphs often exhibit physical symmetries of translations,rotations...Geometric graphs are a special kind of graph with geometric features,which are vital to model many scientific problems.Unlike generic graphs,geometric graphs often exhibit physical symmetries of translations,rotations,and reflections,making them ineffectively processed by current Graph Neural Networks(GNNs).To address this issue,researchers proposed a variety of geometric GNNs equipped with invariant/equivariant properties to better characterize the geometry and topology of geometric graphs.Given the current progress in this field,it is imperative to conduct a comprehensive survey of data structures,models,and applications related to geometric GNNs.In this paper,based on the necessary but concise mathematical preliminaries,we formalize geometric graph as the data structure,on top of which we provide a unified view of existing models from the geometric message passing perspective.Additionally,we summarize the applications as well as the related datasets to facilitate later research for methodology development and experimental evaluation.We also discuss the challenges and future potential directions of geometric GNNs at the end of this survey.展开更多
We propose a novel method that automatically analyzes stroke-related artistic styles of paintings. A set of adaptive interfaces are also developed to connect the style analysis with existing painterly rendering system...We propose a novel method that automatically analyzes stroke-related artistic styles of paintings. A set of adaptive interfaces are also developed to connect the style analysis with existing painterly rendering systems, so that the specific artistic style of a template painting can be effectively transferred to the input photo with minimal effort. Different from conventional texture-synthesis based rendering techniques that focus mainly on texture features, this work extracts, analyzes and simulates high-level style features expressed by artists＇ brush stroke techniques. Through experiments, user studies and comparisons with ground truth, we demonstrate that the proposed style-orientated painting framework can significantly reduce tedious parameter adjustment, and it allows amateur users to efficiently create desired artistic styles simply by specifying a template painting.展开更多
Identification of significant biological relationships or patterns is central to many metagenomic studies.Methods that estimate association networks have been proposed for this purpose;however,they assume that associa...Identification of significant biological relationships or patterns is central to many metagenomic studies.Methods that estimate association networks have been proposed for this purpose;however,they assume that associations are static,neglecting the fact that relationships in a microbial ecosystem may vary with changes in environmental factors(EFs),which can result in inaccurate estimations.Therefore,in this study,we propose a computational model,called the k-Lognormal-Dirichlet-Multinomial(kLDM)model,which estimates multiple association networks that correspond to specific environmental conditions,and simultaneously infers microbe-microbe and EF-microbe associations for each network.The effectiveness of the kLDM model was demonstrated on synthetic data,a colorectal cancer(CRC)dataset,the Tara Oceans dataset,and the American Gut Project dataset.The results revealed that the widely-used Spearman’s rank correlation coefficient method performed much worse than the other methods,indicating the importance of separating samples by environmental conditions.Cancer fecal samples were then compared with cancer-free samples,and the estimation achieved by kLDM exhibited fewer associations among microbes but stronger associations between specific bacteria,especially five CRC-associated operational taxonomic units,indicating gut microbe translocation in cancer patients.Some EF-dependent associations were then found within a marine eukaryotic community.Finally,the gut microbial heterogeneity of inflammatory bowel disease patients was detected.These results demonstrate that kLDM can elucidate the complex associations within microbial ecosystems.The kLDM program,R,and Python scripts,together with all experimental datasets,are accessible at https://github.com/tinglab/kLDM.git.展开更多
基金funded by Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2025R749)Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘Timely and accurate diagnosis of skin diseases is crucial as conventional methods are time-consuming and prone to errors.Traditional trial-and-error approaches often aggregate multiple models without optimization by resulting in suboptimal performance.To address these challenges,we propose a novel Squid Game OptimizationDimension Reduction-based Ensemble(SGO-DRE)method for the precise diagnosis of skin diseases.Our approach begins by selecting pre-trained models named MobileNetV1,DenseNet201,and Xception for robust feature extraction.These models are enhanced with dimension reduction blocks to improve efficiency.To tackle the aggregation problem of various models,we leverage the Squid Game Optimization(SGO)algorithm,which iteratively searches for the optimal weightage set to assign the appropriate weightage to each individual model within the proposed weighted average aggregation ensemble approach.The proposed ensemble method effectively utilizes the strengths of each model.We evaluated the proposed method using an 8-class skin disease dataset,a 6-class MSLD dataset,and a 4-class MSID dataset,achieving accuracies of 98.71%,96.34%,and 93.46%,respectively.Additionally,we employed visual tools like Grad-CAM,ROC curves,and Precision-Recall curves to interpret the decision making of models and assess its performance.These evaluations ensure that the proposed method not only provides robust results but also enhances interpretability and reliability in clinical decision-making.
基金Sponsored by the National High Technology Development Program of China (Grant No. 2002AA142020).
文摘With the increasing enlargement of network scale and the rapid development of network techniques, large numbers of the network applications begin to appear. Packet capture plays an important role as one basic technique used in each field of the network applications. In a high-speed network, the heavy traffic of network transmission challenges the packet capture techniques. This paper does an in-depth analysis on the traditional packet capture mechanisms in Linux, and then measures the performance bottleneck in the process of packet capture. The methods for improving the packet capture performance are presented and an optimized packet capture scheme is also designed and implemented. The test demonstrates that the new packet capture mechanism (Libpacket) can greatly improve the packet capture performance of the network application systems in a high-speed network.
基金This paper is supported by Chunhui pro-gram of MOE(Z2005-1-52015)
文摘This paper shows the harm of harmonic in power system,compares the measures of normal digital filter and wavelet MARto afford reference to the detection and elimination in power system harmonic control.
基金National Natural Science Foundation of China(No.70971020)the Subject of Ministry of Education of Hunan Province,China(No.13C818)+3 种基金the Project of Industrial Science and Technology Support of Hengyang City,Hunan Province,China(No.2013KG63)the Open Project Program of Artificial Intelligence Key Laboratory of Sichuan Province,Sichuan University of Science and Engineering,China(No.2012RYJ03)the Fund Project of Humanities and Social Sciences,Ministry of Education of China(No.13YJCZH147)the Special Fund for Shanghai Colleges' Outstanding Young Teachers' Scientific Research Projects,China(No.ZZGJD12033)
文摘The traveling salesman problem( TSP) is a well-known combinatorial optimization problem as well as an NP-complete problem. A dynamic multi-swarm particle swarm optimization and ant colony optimization( DMPSO-ACO) was presented for TSP.The DMPSO-ACO combined the exploration capabilities of the dynamic multi-swarm particle swarm optimizer( DMPSO) and the stochastic exploitation of the ant colony optimization( ACO) for solving the traveling salesman problem. In the proposed hybrid algorithm,firstly,the dynamic swarms,rapidity of the PSO was used to obtain a series of sub-optimal solutions through certain iterative times for adjusting the initial allocation of pheromone in ACO. Secondly,the positive feedback and high accuracy of the ACO were employed to solving whole problem. Finally,to verify the effectiveness and efficiency of the proposed hybrid algorithm,various scale benchmark problems were tested to demonstrate the potential of the proposed DMPSO-ACO algorithm. The results show that DMPSO-ACO is better in the search precision,convergence property and has strong ability to escape from the local sub-optima when compared with several other peer algorithms.
文摘The scarcity of bandwidth in the radio spectrum has become more vital since the demand for more and more wireless applications has increased. Most of the spectrum bands have been allocated although many studies have shown that these bands are significantly underutilized most of the time. The problem of unavailability of spectrum and inefficiency in its utilization has been smartly addressed by the cognitive radio （CR） technology which is an opportunistic network that senses the environment, observes the network changes, and then uses knowledge gained from the prior interaction with the network to make intelligent decisions by dynamically adapting their transmission characteristics. In this paper, some of the decentralized adaptive medium access control （MAC） protocols for CR networks have been critically analyzed, and a novel adaptive MAC protocol for CR networks, decentralized non-global MAC （DNG-MAC）, has been proposed. The results show the DNG-MAC outperforms other CR-MAC protocols in terms of time and energy efficiency.
文摘A new ontology-based question expansion （OBQE） method is proposed for question similarity calculation in a frequently asked question （FAQ） answering system. Traditional question similarity calculation methods use ＂word＂ to compose question vector, that the semantic relations between words are ignored. OBQE takes the relation as an important part. The process of the new system is：① to build two-layered domain ontology referring to WordNet and domain corpse;② to expand question trunks into domain cases;③ to use domain case composed vector to calculate question similarity. The experimental result shows that the performance of question similarity calculation with OBQE is being improved.
文摘A valid method of virtual scene depth calculating is put forward. In this method cameras rotate in three different viewpoints in the plane and we calculate the depth of panorama using three stitching cylinder panoramas. In the investigation, the column of panorama is regarded as a slot image. Using the conic intersected by the epipolar plane and the cylinder, we can obtain the pel-pendicularity disparity. In order to obtain dense correspondence fast and accurately, a new method of obtaining horizontal disparity using depth continuity is also put forward. It converts the problem of panorama dense correspondence to the problem of searching points in the conic. The occlusion problem is dealt with using three cylinders in the depth calculation. It is verified that this method is convenient, useful and efficient in calculating the depth of a virtual scene.
基金Sponsored by the National 863 Plan (Grant No.2002AA1Z2101)the National Tenth Five-Year Research Plan(Grant No. 41316.1.2).
文摘The mode of mobile computing originated from distributed computing and it has the un-idempotent operation property, therefore the deadlock detection algorithm designed for mobile computing systems will face challenges with regard to correctness and high efficiency. This paper attempts a fundamental study of deadlock detection for the AND model of mobile computing systems. First, the existing deadlock detection algorithms for distributed systems are classified into the resource node dependent (RD) and the resource node independent (RI) categories, and their corresponding weaknesses are discussed. Afterwards a new RI algorithm based on the AND model of mobile computing system is presented. The novelties of our algorithm are that: 1) the blocked nodes inform their predecessors and successors simultaneously; 2) the detection messages (agents) hold the predecessors information of their originator; 3) no agent is stored midway. Additionally, the quit-inform scheme is introduced to treat the excessive victim quitting problem raised by the overlapped cycles. By these methods the proposed algorithm can detect a cycle of size n within n-2 steps and with (n^2-n-2)/2 agents. The performance of our algorithm is compared with the most competitive RD and RI algorithms for distributed systems on a mobile agent simulation platform. Experiment results point out that our algorithm outperforms the two algorithms under the vast majority of resource configurations and concurrent workloads. The correctness of the proposed algorithm is formally proven by the invariant verification technique.
基金supported by the National Natural Science Foundation of China under Grant No.51679058the China Higher Specialized Research Fund(Ph.D.supervisor category) under Grant No.20132304110018
文摘The quantum theory application is a hot research area in recent years,especially the theory of quantum mechanics.In this paper,we focus on the research of image segmentation based on quantum mechanics.Firstly,the theory of quantum mechanics is introduced;afterwards,a review of image segmentation methods based on quantum mechanics is presented;and finally,the characteristics about the quantum mechanics applied to image processing are concluded.Two main research topics are discussed in this paper.One is to emphasize that quantum mechanics can be applied in different research areas,such as image segmentation,and the second is to conclude some methods in image segmentation and give some suggestions for possible novel methods by applying quantum mechanics theory.As a summary,this is a review paper which presents some methods based on the feasible theory in quantum mechanics aiming at achieving a better performance in image segmentation.
文摘TikTok is one of the world’s most attractive and fastest-growing social media platforms.It has more than 2.6 billion downloads globally and over 100 million users in the US.The“secret weapon”is its unique methodology of discovering and delivering content.On the“For You”feed,TikTok mixes videos of both internet celebrities and newcomers,encourages high-quality creative content according to page views,and promotes new bloggers to share videos with users.The unique aspect is that anyone has the opportunity to spring into fame on the feed.Through TikTok’s recom-mendation algorithm,videos are continuously recommended to users with similar interests or attributes as video bloggers,thus allowing high-quality creative content to be disseminated quickly.The TikTok recommendation algorithm does not take video blogger’s fanbase or popularity into major consideration.In addition to the candidate video title,audio,and tags,the algorithm integrates the content of the user’s uploaded videos,and the categories of videos liked by the user.TikTok’s skills at enhancing user loyalty are impressive,which can not only accurately recommend videos of interest to users,but also assist them in expanding into new intersecting areas.
基金This work was supported by the National Key Research and Development Program of China(2018YFC0831700)the National Natural Science Foundation of China(Grant Nos.61732008,61532011)Beijing Academy of Artificial Intelligence(BAAI)and Tsinghua University Guoqiang Research Institute.
文摘Relevance estimation is one of the core concerns of information retrieval(IR)studies.Although existing retrieval models gained much success in both deepening our understanding of information seeking behavior and building effective retrieval systems,we have to admit that the models work in a rather different manner from how humans make relevance judgments.Users’information seeking behaviors involve complex cognitive processes,however,the majority of these behavior patterns are not considered in existing retrieval models.To bridge the gap between practical user behavior and retrieval model,it is essential to systematically investigate user cognitive behavior during relevance judgement and incorporate these heuristics into retrieval models.In this paper,we aim to formally define a set of basic user reading heuristics during relevance judgement and investigate their corresponding modeling strategies in retrieval models.Further experiments are conducted to evaluate the effectiveness of different reading heuristics for improving ranking performance.Based on a large-scale Web search dataset,we find that most reading heuristics can improve the performance of retrieval model and establish guidelines for improving the design of retrieval models with human-inspired heuristics.Our study sheds light on building retrieval model from the perspective of cognitive behavior.
基金supported by the following projects:The National Natural Science Foundation of China(Grant Nos.62376276 and 62172422)Beijing Nova Program(Grant No.20230484278)+1 种基金the Fundamental Research Funds for the Central Universities,and the Research Funds of Renmin University of China(Grant No.23XNKJ19)Tencent AI Lab Rhino-Bird Focused Research Program.
文摘Geometric graphs are a special kind of graph with geometric features,which are vital to model many scientific problems.Unlike generic graphs,geometric graphs often exhibit physical symmetries of translations,rotations,and reflections,making them ineffectively processed by current Graph Neural Networks(GNNs).To address this issue,researchers proposed a variety of geometric GNNs equipped with invariant/equivariant properties to better characterize the geometry and topology of geometric graphs.Given the current progress in this field,it is imperative to conduct a comprehensive survey of data structures,models,and applications related to geometric GNNs.In this paper,based on the necessary but concise mathematical preliminaries,we formalize geometric graph as the data structure,on top of which we provide a unified view of existing models from the geometric message passing perspective.Additionally,we summarize the applications as well as the related datasets to facilitate later research for methodology development and experimental evaluation.We also discuss the challenges and future potential directions of geometric GNNs at the end of this survey.
基金supported by Fok Ying-Tong Education Foundation of China under Grant No. 131065the International Joint Project from the Royal Society of UK under Grant No. JP100987
文摘We propose a novel method that automatically analyzes stroke-related artistic styles of paintings. A set of adaptive interfaces are also developed to connect the style analysis with existing painterly rendering systems, so that the specific artistic style of a template painting can be effectively transferred to the input photo with minimal effort. Different from conventional texture-synthesis based rendering techniques that focus mainly on texture features, this work extracts, analyzes and simulates high-level style features expressed by artists＇ brush stroke techniques. Through experiments, user studies and comparisons with ground truth, we demonstrate that the proposed style-orientated painting framework can significantly reduce tedious parameter adjustment, and it allows amateur users to efficiently create desired artistic styles simply by specifying a template painting.
基金supported by the National Natural Science Foundation of China(Grant Nos.61872218,61673241,and 61721003)the Tsinghua-Fuzhou Institute Research ProgramBeijing National Research Center for Information Science and Technology(BNRist),China。
文摘Identification of significant biological relationships or patterns is central to many metagenomic studies.Methods that estimate association networks have been proposed for this purpose;however,they assume that associations are static,neglecting the fact that relationships in a microbial ecosystem may vary with changes in environmental factors(EFs),which can result in inaccurate estimations.Therefore,in this study,we propose a computational model,called the k-Lognormal-Dirichlet-Multinomial(kLDM)model,which estimates multiple association networks that correspond to specific environmental conditions,and simultaneously infers microbe-microbe and EF-microbe associations for each network.The effectiveness of the kLDM model was demonstrated on synthetic data,a colorectal cancer(CRC)dataset,the Tara Oceans dataset,and the American Gut Project dataset.The results revealed that the widely-used Spearman’s rank correlation coefficient method performed much worse than the other methods,indicating the importance of separating samples by environmental conditions.Cancer fecal samples were then compared with cancer-free samples,and the estimation achieved by kLDM exhibited fewer associations among microbes but stronger associations between specific bacteria,especially five CRC-associated operational taxonomic units,indicating gut microbe translocation in cancer patients.Some EF-dependent associations were then found within a marine eukaryotic community.Finally,the gut microbial heterogeneity of inflammatory bowel disease patients was detected.These results demonstrate that kLDM can elucidate the complex associations within microbial ecosystems.The kLDM program,R,and Python scripts,together with all experimental datasets,are accessible at https://github.com/tinglab/kLDM.git.
