检索结果-内蒙古大学图书馆

2019 TREC Video Retrieval Evaluation, TRECVID 2019

作者： Li, Ya Chen, Guanyu Cheng, Xiangqian Chen, Chong Xu, Shaoqiang Li, Xinyu Xiang, Xuanlu Zhao, Yanyun Zhao, Zhicheng Su, Fei Multimedia Communication and Pattern Recognition Lab Beijing Key Laboratory of Network System Network Culture Beijing University of Posts and Telecommunications China

In this paper, we describe BUPT-MCPRL systems and evaluation results for TRECVID 2019 [14]. We join two tasks: activities in extended video and instance search. Activities in Extended Video (ActEV): p_baseline_2: the baseline of our method. In this submission, we use 3D-Faster-RCNN with Focal Loss to generate activity tubes, then we use tracking algorithm to link the tubes and build efficient temporal localization systems to detect events (activity instances). p_baseline_14: add OTW dataset to train 3D-Faster-RCNN. p_baseline_22: split long events. p_baseline_23: fuse similar events. Instance Search (INS): F_M_E_E_BUPT_MCPRL_1: retrieve persons and using pose-based action detection F_M_E_E_BUPT_MCPRL_2: retrieve persons and detect their actions F_P_E_E_BUPT_MCPRL_3: the same method as in F_M_E_E_BUPT_MCPRL_2I_M_E_E_BUPT_MCPRL_4: first retrieve target persons and detect their actions with partial correction manually. Copyright © TRECVID *** rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Ontology-Semantic Alignment On Contrastive Video-Language Model for Multimodel Video Retrieval Task 7

Ontology-Semantic Alignment On Contrastive Video-Language Mo...

引用

7th International Conference on Computer Science and Artificial Intelligence, CSAI 2023

作者： Zhou, Yifan Ding, Yizhou Dong, Yuwu He, Hao Shanghai Jiao Tong University State Key Lab of Advanced Optical Communication System and Network Shanghai China

ISBN: (纸本)9798400708688

Contrastive Learning-based models have shown impressive performance in text-image retrieval tasks. However, when applied in video retrieval, traditional contrastive learning strategies have faced challenges in achieving satisfactory results due to redundancy of video contents. We discern several potential reasons: (1)Current methodologies sometimes overlook the significant information imbalance between videos and query text, specifically neglecting the in-depth textual representation of the content within the videos. (2) Current video matching methodologies typically focus on cross-model alignment at general entity similarity level, without specific consideration for how entity pair preferences and similarity properties affect the task at hand. (3) Previous vectorized retrieval based on video content features have been somewhat flawed. They primarily focused on aligning overall features without having an video content tags feature for meaningful feature discrimination. Considering the shortcomings identified in the mentioned three aspects, we propose an ontology semantic labels augments retrieval model and introduce a method to integrate video ontology semantic labels into the contrastive learning framework. In particular, we have developed ontology semantic descriptions about entities encompassing both human figures and textual elements within the videos. Subsequently, we conducted training and testing on the CMIVQA dataset to assess the performance of our approach. The experimental results show that employing fine-grained ontology labels as sample pairs for contrastive learning leads to an increased level of precision in video retrieval tasks. © 2023 ACM.

关键词： Ontology

来源：评论

学校读者我要写书评

暂无评论

LDSeq: Latent Diffusion Models for Sequence to Sequence Text Generation 7

LDSeq: Latent Diffusion Models for Sequence to Sequence Text...

引用

7th International Conference on Computer Science and Artificial Intelligence, CSAI 2023

作者： Ding, Yizhou Tian, Jidong Mei, Shanxing Zhou, Yifan Dong, Yuwu He, Hao Hu, Weisheng State Key Lab of Advanced Optical Communication System and Network Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9798400708688

Diffusion models have demonstrated remarkable success in generating continuous data, such as images and audios. Previous studies on text generation employing continuous diffusion models have revealed the potential of the diffusion framework. However, challenges like embedding collapse persist, limiting the overall generation performance. In this paper we introduce LDSeq, a latent diffusion framework employing a two-stage training procedure for sequence-to-sequence text generation. In the proposed framework, we first train a Variational Auto-Encoder (VAE) on downstream datasets to compress the target text of samples into a continuous latent space, and then we train a conditional latent diffusion model in the fixed continuous latent space, where the latent vectors are iteratively sampled conditioned on the input source text. The disjoint training stages prevent the collapse of diffusion space. Experimental results on paraphrase generation and text summarization datasets show that LDSeq achieves comparable or superior performance in comparison to AR and NAR baselines while requiring lower training cost. Furthermore, We discuss some potential future directions for enhancing diffusion models in the text generation domain. © 2023 ACM.

关键词： Diffusion

来源：评论

学校读者我要写书评

暂无评论

The design and implementation of distributed resource manager in optical grid networks

The design and implementation of distributed resource manage...

引用

2009 Asia communications and Photonics Conference and Exhibition, ACP 2009

作者： Chen, Siwei Hu, Weisheng Guo, Wei Jin, Yaohui State Key Lab. of Advanced Optical Communication System and Network Shanghai Jiao Tong University Shanghai China

Effective management of large amount of heterogeneous resources in optical grid is critical to the performance of the grid. This article proposes a distributed system for an integrated management of both application a... 详细信息

ISBN: (纸本)9781557528773

关键词： Photonics

来源：评论

学校读者我要写书评

暂无评论

On the Feasibility of Reconfigurable Photonic network on Chip

On the Feasibility of Reconfigurable Photonic Network on Chi...

引用

Conference on Lasers and Electro-Optics/Pacific Rim, CLEOPR 2009

作者： Jin, Yaohui Tang, Jianxiong Chang, Zhijuan Gao, Yu Hu, Weisheng State Key Lab of Advanced Optical Communication System and Network Shanghai Jiao Tong university Shanghai 200240 China

Reconfigurable photonic network formulti-processor on chip communication is considered. Withdifferent application pattern, this architecture can significantlyimprove the latency performance of multi-processor chips. &... 详细信息

ISBN: (纸本)9781424438303

关键词： network-on-chip

来源：评论

学校读者我要写书评

暂无评论

RETRACTED ARTICLE: An improved tabu optimization approach for workflow scheduling

RETRACTED ARTICLE: An improved tabu optimization approach fo...

引用

作者： Xu, Zhan Wu, Gang Wang, Jin Lin Liu, Lei Key Lab of Network Communication System and Control USTC Chinese Academy of Sciences Hefei China National Network New Media Engineering Research Center Institute of Acoustics Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781424455386

Directed Acyclic Graph (DAG) is a nature way of workflow description. One of the most important and challenging problems in the DAG workflow field is the QoS-constrained scheduling problem with the aim to minimize the cost. This paper proposes a new algorithm to tackle this problem. The algorithm proposed is based on tabu search and make some improvements. Subject to the deadline of the workflow, this algorithm has a lower execution cost than other existing algorithms. Simulation results prove the effectiveness and feasibility of the algorithm. © 2010 IEEE.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

DPNet: Dynamic Pooling network for Accurate and Efficient Size-Aware Tiny Object Detection

引用

IEEE Internet of Things Journal 2025年

作者： Gong, Luqi Chen, Haotian Chen, Yikun Yao, Tianliang Li, Chao Zhao, Shuai Han, Guangjie Beijing University of Posts and Telecommunications School of Computer Science Beijing China Zhejiang Lab Research Center for Space Computing System Hangzhou China Southwest Jiaotong University SWJTU-LEEDS Joint School Chengdu China LTD Guangdong Zhiyun City construction Technology Co Zhuhai China Tongji University College of Electronic and Information Engineering Department of Control Science and Engineering Shanghai China Hohai University Key Laboratory of Maritime Intelligent Network Information Technology Ministry of Education China

In unmanned aerial systems, especially in complex environments, accurately detecting tiny objects is crucial. Resizing images is a common strategy to improve detection accuracy, particularly for small objects. However, simply enlarging images significantly increases computational costs and the number of negative samples, severely degrading detection performance and limiting its applicability. This paper proposes a Dynamic Pooling network (DPNet) for tiny object detection to mitigate these issues. DPNet employs a flexible down-sampling strategy by introducing a factor (df) to relax the fixed down-sampling process of the feature map to an adjustable one. Furthermore, we design a lightweight predictor to predict df for each input image, which will be used to decrease the resolution of feature map in backbone. Thus, we achieve input-aware down-sampling. We design an Adaptive Normalization Module (ANM) to make a unified detector well compatible with different dfs. At the same time, we also design a guidance loss to supervise the predictor’s training. DPNet realizes the dynamic allocation of computing resources to trade off detection accuracy and efficiency through this. Experiments on the TinyCOCO and TinyPerson datasets show that our DPNet can save over 35% and 25% GFLOPs, respectively, while maintaining comparable detection *** code will be made publicly available. © 2014 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Deflection routing in multi-channel photonic network on chip architecture

Deflection routing in multi-channel photonic network on chip...

引用

network Architectures, Management, and Applications VII

作者： Tang, Jianxiong Jin, Yaohui Chang, Zhijuan State Key Lab of Advanced Optical Communication System and Network Shanghai Jiao Tong University Shanghai 200240 China

ISBN: (纸本)9780819480354

Ultralow-latency and less power consumption have become necessary in multi-processor interconnection network on chip, photonic interconnection as a solution to meet above requirement, provides high performance interconnection on chip. But the photonic network on chip architecture design and performance is limited because photonic interconnection hasn't buffer, photonic network architecture must be designed to relieve this limitation. In this paper, we present a multi-channel photonic network on chip architecture employing deflection routing, optical data packets can inject/eject from processor core by four channels at the same time. Simulation result shows this network architecture has 60% latency decrease compared to generic photonic network on chip, and the photonic network architecture is only consume 7% power of the electronic interconnection network on chip with the same scale. © 2009 SPIE-OSA-IEEE.

关键词： network-on-chip

来源：评论

学校读者我要写书评

暂无评论

Dynamic overlay routing based on active probing measurements: An emulation study

Dynamic overlay routing based on active probing measurements...

引用

network Architectures, Management, and Applications VII

作者： Zhang, Xinxin Ye, Wei Jin, Yaohui State Key Lab of Advanced Optical Communication System and Network Shanghai Jiao Tong Univ. Shanghai China

ISBN: (纸本)9780819480354

Recent studies have shown that path diversity provided by overlay networks brings about possibilities to choose alternate optimal path against default one. In the present paper, we present a novel routing mechanism which chooses path based on a composite metric, whose comprising individual metrics are determined by active probing measurements including delay, jitter and bandwidth on the overlay network in a real-time fashion. The applicability of this mechanism is examined under overlay emulation environment, which is supported by the performance data collected from a real enterprise overlay network. © 2009 SPIE-OSA-IEEE.

关键词： network routing

来源：评论

学校读者我要写书评

暂无评论

A high-speed real-time HTTP performance measurement architecture based on network processor

A high-speed real-time HTTP performance measurement architec...

引用

2011 International Conference on ICT Convergence, ICTC 2011

作者： Yang, Xin Chen, Xiaming Jin, Yaohui State Key Lab. of Advanced Optical Communication System and Network Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9781457712678

In this paper, we propose an architecture for highspeed real-time passive HTTP traffic performance measurement. High-speed real-time measurement is hard to realize due to the requirement of high capturing and processing power. We introduce network processor to solve this problem. Then we deploy the system at the Internet access link of a campus network, and demonstrate the processing capability and feasibility of this architecture through the measurement results. © 2011 IEEE.

关键词： HTTP

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：