检索结果-南通市图书馆

A lightweight convolutional neural network for large-scale Chinese image caption

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Optoelectronics Letters 2021年第6期17卷 361-366页

作者：赵德新杨瑞雪郭淑涛 tianjin key laboratory of intelligence computing and novel software technology Tianjin University of TechnologyTianjin 300384China

Image caption is a high-level task in the area of image understanding,in which most of the models adopt a convolutional neural network(CNN)to extract image features assigning a recurrent neural network(RNN)to generate sentences.Researchers tend to design complex networks with deeper layers to improve the performance of feature extraction in recent years.Increasing the size of the network could obtain features of high quality,but it is not an efficient way in terms of computational cost.A large number of parameters brought by CNN makes the research difficult to apply in human daily life.In order to reduce the information loss of the convolutional process with less cost,we propose a lightweight convolutional neural network,named as Bifurcate-CNN(B-CNN).Furthermore,recent works are devoted to generating captions in English,in this paper,we develop an image caption model that generates descriptions in Chinese.Compared with Inception-v3,the depth of our model is shallower with fewer parameters,and the computational cost is lower.Evaluated on the AI CHALLENGER dataset,we prove that our model can enhance the performance,improving BLEU-4 from 46.1 to 49.9 and CIDEr from 142.5 to 156.6 respectively.

关键词： neural network image

Solid-state green-light-emitting upconversion coherent random laser in macroporous LiNbO_3:Er

在线全文

学校读者我要写书评

暂无评论

Chinese Optics Letters 2018年第1期16卷 62-65页

作者：石凡赵萌贾晨赵宇峰陈胜勇 key laboratory of Computer Vision and System of Ministry of Education Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology Tianjin University of Technology

A solid-state green-light-emitting upconversion coherent random laser was realized by pumping macroporous erbium-doped lithium niobate with a 980 nm laser. The lasing threshold was determined to be about 40 k W∕cm~2.Above the threshold, the emission intensity increased sharply with the increasing pump intensity. Moreover, a narrow multi-peaks structure was observed in the green-light-emission band, and the positions of lasing lines were various at different angles. The results were the direct evidences of coherent random lasing emission from macroporous erbium-doped lithium niobate. These phenomena were attributed to the coexistence of upconversion emission and a multiple scattering feedback mechanism.

关键词： Solid-state green-light-emitting upconversion coherent random laser in macroporous LiNbO3:Er LN ETU

CKGSE:A Prototype Search Engine for Chinese Knowledge Graphs

在线全文

学校读者我要写书评

暂无评论

Data intelligence 2022年第1期4卷 41-65页

作者： Xiaxia Wang Tengteng Lin Weiqing Luo Gong Cheng Yuzhong Qu State key laboratory for novel software technology Nanjing UniversityNanjing 210023China

Nowadays,with increasing open knowledge graphs(KGs)being published on the Web,users depend on open data portals and search engines to find KGs.However,existing systems provide search services and present results with only metadata while ignoring the contents of KGs,i.e.,triples.It brings difficulty for users’comprehension and relevance judgement.To overcome the limitation of metadata,in this paper we propose a content-based search engine for open KGs named CKGSE.Our system provides keyword search,KG snippet generation,KG profiling and browsing,all based on KGs’detailed,informative contents rather than their brief,limited metadata.To evaluate its usability,we implement a prototype with Chinese KGs crawled from Open *** and report some preliminary results and findings.

关键词： Knowledge graph Search engine Snippet generation Dataset profiling Browsing

Semantic image segmentation with fused CNN features

在线全文

学校读者我要写书评

暂无评论

Optoelectronics Letters 2017年第5期13卷 381-385页

作者：耿慧强张桦薛彦兵周冕徐光平高赞 key laboratory of Computer Vision and System of Ministry of Education Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology Tianjin University of Technology

Semantic image segmentation is a task to predict a category label for every image pixel. The key challenge of it is to design a strong feature representation. In this paper, we fuse the hierarchical convolutional neural network(CNN) features and the region-based features as the feature representation. The hierarchical features contain more global information, while the region-based features contain more local information. The combination of these two kinds of features significantly enhances the feature representation. Then the fused features are used to train a softmax classifier to produce per-pixel label assignment probability. And a fully connected conditional random field(CRF) is used as a post-processing method to improve the labeling consistency. We conduct experiments on SIFT flow dataset. The pixel accuracy and class accuracy are 84.4% and 34.86%, respectively.

关键词：图像分割电视新闻特征表语义有线美国图像像素卷积神经网络

Efficient Multiagent Policy Optimization Based on Weighted Estimators in Stochastic Cooperative Environments

在线全文

学校读者我要写书评

暂无评论

Journal of Computer Science & technology 2020年第2期35卷 268-280页

作者： Yan Zheng Jian-Ye Hao Zong-Zhang Zhang Zhao-Peng Meng Xiao-Tian Hao College of intelligence and computing Tianjin UniversityTianjin 300350China National key laboratory for novel software technology Nanjing UniversityNanjing 210023China

Multiagent deep reinforcement learning (MA-DRL) has received increasingly wide attention. Most of the existing MA-DRL algorithms, however, are still inefficient when faced with the non-stationarity due to agents changing behavior consistently in stochastic environments. This paper extends the weighted double estimator to multiagent domains and proposes an MA-DRL framework, named Weighted Double Deep Q-Network (WDDQN). By leveraging the weighted double estimator and the deep neural network, WDDQN can not only reduce the bias effectively but also handle scenarios with raw visual inputs. To achieve efficient cooperation in multiagent domains, we introduce a lenient reward network and scheduled replay strategy. Empirical results show that WDDQN outperforms an existing DRL algorithm (double DQN) and an MA-DRL algorithm (lenient Q-learning) regarding the averaged reward and the convergence speed and is more likely to converge to the Pareto-optimal Nash equilibrium in stochastic cooperative environments.

关键词： deep reinforcement learning multiagent system weighted double estimator lenient reinforcement learning cooperative Markov game

Behavior of Collective Cooperation Yielded by Two Update Rules in Social Dilemmas:Combining Fermi and Moran Rules

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Communications in Theoretical Physics 2012年第9期58卷 343-348页

作者：夏承遗王磊王娟王劲松 key laboratory of Computer Vision and System(Ministry of Education)and tianjin key laboratory of intelligence computing and novel software technology Tianjin University of Technology School of Automation Tianjin University of Technology

We combine the Fermi and Moran update rules in the spatial prisoner's dilemma and snowdrift games to investigate the behavior of collective cooperation among agents on the regular lattice.Large-scale simulations indicate that,compared to the model with only one update rule,the cooperation behavior exhibits the richer phenomena,and the role of update dynamics should be paid more attention in the evolutionary game theory.Meanwhile,we also observe that the introduction of Moran rule,which needs to consider all neighbor's information,can markedly promote the aggregate cooperation level,that is,randomly selecting the neighbor proportional to its payoff to imitate will facilitate the cooperation among agents.Current results will contribute to further understand the cooperation dynamics and evolutionary behaviors within many biological,economic and social systems.

关键词：社会制度合作行为集体费米进化博弈论囚徒困境随机选择

AE-TPGG:a novel autoencoder-based approach for single-cell RNA-seq data imputation and dimensionality reduction

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2023年第3期17卷 217-234页

作者： Shuchang ZHAO Li ZHANG Xuejun LIU MIIT key laboratory of Pattern Analysis and Machine intelligence College of Computer Science and TechnologyNanjing University of Aeronautics and AstronauticsNanjing 211106China Collaborative Innovation Center of novel software technology and Industrialization Nanjing 210023China College of Computer Science and technology Nanjing Forestry UniversityNanjing 210037China

Single-cell RNA sequencing(scRNA-seq)technology has become an effective tool for high-throughout transcriptomic study,which circumvents the averaging artifacts corresponding to bulk RNA-seq technology,yielding new perspectives on the cellular diversity of potential superficially homogeneous populations.Although various sequencing techniques have decreased the amplification bias and improved capture efficiency caused by the low amount of starting material,the technical noise and biological variation are inevitably introduced into experimental process,resulting in high dropout events,which greatly hinder the downstream analysis.Considering the bimodal expression pattern and the right-skewed characteristic existed in normalized scRNA-seq data,we propose a customized autoencoder based on a twopart-generalized-gamma distribution(AE-TPGG)for scRNAseq data analysis,which takes mixed discrete-continuous random variables of scRNA-seq data into account using a twopart model and utilizes the generalized gamma(GG)distribution,for fitting the positive and right-skewed continuous data.The adopted autoencoder enables AE-TPGG to captures the inherent relationship between genes.In addition to the ability of achieving low-dimensional representation,the AETPGG model also provides a denoised imputation according to statistical characteristic of gene expression.Results on real datasets demonstrate that our proposed model is competitive to current imputation methods and ameliorates a diverse set of typical scRNA-seq data analyses.

关键词： scRNA-seq autoencoder TPGG data imputation dimensionality reduction

Evolution of Cooperation in Public Goods Games

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Communications in Theoretical Physics 2011年第10期56卷 638-644页

作者：夏承遗张娟娟王祎玲王劲松 key laboratory of Computer Vision and System(tianjin University of technology) Ministry of Education key laboratory of intelligence computing and novel software technology Tianjin University of Technology College of Life Science Shanxi Normal University

We investigate the evolution of cooperation with evolutionary public goods games based on finite populations, where four pure strategies:cooperators,defectors,punishers and loners who are unwilling to participate are considered.By adopting approximate best response dynamics,we show that the magnitude of rationality not only quantitatively explains the experiment results in[Nature(London) 425(2003) 390],but also it will heavily influence the evolution of cooperation.Compared with previous results of infinite populations,which result in two equilibriums, we show that there merely exists a special equilibrium and the relevant high value of bounded rationality will sustain cooperation.In addition,we characterize that loner’s payoff plays an active role in the maintenance of cooperation, which will only be warranted for the low and moderate values of loner’s payoff.It thus indicates the effects of rationality and loner’s payoff will influence the cooperation.Finally,we highlight the important result that the introduction of voluntary participation and punishment will facilitate cooperation greatly.

关键词：合作进化公共物品游戏有限理性反应动力学高附加值纯策略

Effect of Growing Size of Interaction Neighbors on the Evolution of Cooperation in Spatial Snowdrift Game

在线全文

学校读者我要写书评

暂无评论

Communications in Theoretical Physics 2012年第4期57卷 541-546页

作者：张娟娟王娟孙世温王莉王震夏承遗 key laboratory of Computer Vision and System(Ministry of Education)and tianjin key laboratory of intelligence computing and novel software technology Tianjin University of Technology School of Automation Tianjin University of Technology School of software technology Dalian University of Technology

In this paper,we study the influence of the size of interaction neighbors(k) on the evolution of cooperation in the spatial snowdrift game.At first,we consider the effects of noise K and cost-to-benefit ratio r,the simulation results indicate that the evolution of cooperation depends on the combined action of noise and cost-to-benefit ratio.For a lower r,the cooperators are multitudinous and the cooperation frequency ultimately increases to 1 as the increase of noise.However,for a higher r,the defectors account for the majority of the game and dominate the game if the noise is large enough.Then we mainly investigate how k influences the evolution of cooperation by varying the noise in detail.We find that the frequency of cooperators is closely related to the size of neighborhood and cost-to-benefit ratio r.In the case of lower r,the augmentation of k plays no positive role in promoting the cooperation as compared with that of k = 4,while for higher r the cooperation is improved for a growing size of neighborhood.At last,based on the above discussions,we explore the cluster-forming mechanism among the cooperators.The current results are beneficial to further understand the evolution of cooperation in many natural,social and biological systems.

关键词：合作伙伴演变空间邻居博弈成本效益仿真结果联合行动