检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Qin, Ran Ma, Haoxiang Gao, Boyang Huang, Di State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Beijing China School of Computer Science and Technology Harbin Institute of Technology Harbin China

Planar grasp detection is one of the most fundamental tasks to robotic manipulation, and the recent progress of consumer-grade RGB-D sensors enables delivering more comprehensive features from both the texture and shape modalities. However, depth maps are generally of a relatively lower quality with much stronger noise compared to RGB images, making it challenging to acquire grasp depth and fuse multi-modal clues. To address the two issues, this paper proposes a novel learning based approach to RGB-D grasp detection, namely Depth Guided Cross-modal Attention Network (DGCAN). To better leverage the geometry information recorded in the depth channel, a complete 6-dimensional rectangle representation is adopted with the grasp depth dedicatedly considered in addition to those defined in the common 5-dimensional one. The prediction of the extra grasp depth substantially strengthens feature learning, thereby leading to more accurate results. Moreover, to reduce the negative impact caused by the discrepancy of data quality in two modalities, a Local Cross-modal Attention (LCA) module is designed, where the depth features are refined according to cross-modal relations and concatenated to the RGB ones for more sufficient fusion. Extensive simulation and physical evaluations are conducted and the experimental results highlight the superiority of the proposed approach. Copyright © 2023, The Authors. All rights reserved.

关键词： Textures

来源：评论

学校读者我要写书评

暂无评论

Zone Based Writing Optimization in User Space 23

Zone Based Writing Optimization in User Space

引用

23rd IEEE International Conference on High Performance Computing and Communications, 7th IEEE International Conference on Data science and Systems, 19th IEEE International Conference on Smart City and 7th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Applications, HPCC-DSS-SmartCity-DependSys 2021

作者： Shen, Runnan Liu, Yuting Zhang, Rui Zhu, Jinbin Liu, Zhonglin Xiao, Limin School of Computer Science and Engineering Beihang University Beijing100191 China North China Institute of Computing Technology Beijing China Beihang University State Key Laboratory of Software Development Environment Beijing100191 China

ISBN: (纸本)9781665494571

Write operation of the latest storage media such as NAND SSDs has certain restrictions. Data can only be written in an additional write mode. When erasing, some valid data still needs to be written again, which will cause write amplification. Researches about write amplification focused on conventional NVMe protocol and did not consider some forward features. In response to this problem, this paper designs a user-mode write operation optimization algorithm based on the dynamic adjustment of zone space. When the block is erased, the impact caused by write amplification can be significantly resolved, and operations of multiple tenants can be effectively isolated. The Zone Based Algorithm will implement a mapping algorithm based on zone space, and select a suitable GC to maximize the performance. We also designed and implemented a variety of mapping algorithms and GC algorithms to improve the processing efficiency of IO requests. Experiments based on FEMU show at least 25% improvement when increasing data of different life span. © 2021 IEEE.

关键词： Protocols Smart cities Heuristic algorithms Clustering algorithms Writing Media Jitter

来源：评论

学校读者我要写书评

暂无评论

Prune and Merge: Efficient Token Compression for Vision Transformer with Spatial Information Preserved

引用

IEEE Transactions on Multimedia 2025年

作者： Mao, Junzhu Shen, Yang Guo, Jinyang Yao, Yazhou Hua, Xiansheng Shen, Hengtao Nanjing University of Science and Technology School of Computer Science and Engineering Nanjing210094 China Beihang University State Key Laboratory of Software Development Environment Institute of Artificial Intelligence Beijing100191 China Terminus Group Beijing100027 China University of Electronic Science and Technology of China School of Computer Science and Engineering Chengdu611731 China

Token compression is essential for reducing the computational and memory requirements of transformer models, enabling their deployment in resource-constrained environments. In this work, we propose an efficient and hardware-compatible token compression method called Prune and Merge. Our approach integrates token pruning and merging operations within transformer models to achieve layer-wise token compression. By introducing trainable merge and reconstruct matrices and utilizing shortcut connections, we efficiently merge tokens while preserving important information and enabling the restoration of pruned tokens. Additionally, we introduce a novel gradient-weighted attention scoring mechanism that computes token importance scores during the training phase, eliminating the need for separate computations during inference and enhancing compression efficiency. We also leverage gradient information to capture the global impact of tokens and automatically identify optimal compression structures. Extensive experiments on the ImageNet-1k and ADE20K datasets validate the effectiveness of our approach, achieving significant speed-ups with minimal accuracy degradation compared to state-of-the-art methods. For instance, on DeiT-Small, we achieve a 1.64× speed-up with only a 0.2% drop in accuracy on ImageNet-1k. Moreover, by compressing segmenter models and comparing with existing methods, we demonstrate the superior performance of our approach in terms of efficiency and effectiveness. Code and models have been made availab.e at https://***/NUST-Machine-Intelligence-lab.ratory/prune_and_merge. © 1999-2012 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

User-level failure detection and auto-recovery of parallel programs in HPC systems

引用

Frontiers of computer science 2021年第6期15卷 31-42页

作者： Guozhen ZHANG Yi LIU Hailong YANG Jun XU Depei QIAN State Key Laboratory of Software Development Environment Beijing 100191China Sino-German Joint Software Institute Beihang UniversityBeijing 100191China School of Computer Science and Engineering Beihang UniversityBeijing 100191China Science and Technology on Space System Simulation Laboratory Beijing Simulation Center Beijing 100854China

As the mean-time-between-failures(MTBF)continues to decline with the increasing number of components on large-scale high performance computing(HPC)systems,program failures might occur during the execution period with high *** successful execution of the HPC programs has become an issue that the unprivileged users should be *** the user perspective,if the program failure cannot be detected and handled in time,it would waste resources and delay the progress of program ***,the unprivileged users are unable to perform program state checking due to execution control by the job management system as well as the limited ***,automated tools for supporting user-level failure detection and autorecovery of parallel programs in HPC systems are *** paper proposes an innovative method for the unprivileged user to achieve failure detection of job execution and automatic resubmission of failed *** state checker in our method is encapsulated as an independent job to reduce interference with the user *** addition,we propose a dual-checker mechanism to improve the robustness of our *** implement the proposed method as a tool named automatic re-launcher(ARL)and evaluate it on the Tianhe-2 *** results show that ARL can detect the execution failures effectively on Tianhe-2 *** addition,the communication and performance overhead caused by ARL is *** good scalab.lity of ARL makes it applicable for large-scale HPC systems.

关键词： high performance computing parallel program failure detection failure auto-recovery

来源：评论

学校读者我要写书评

暂无评论

Zero-Shot Scene Graph Generation via Triplet Calibration and Reduction

arXiv

引用

arXiv 2023年

作者： Li, Jiankai Wang, Yunhong Li, Weixin State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Shanghai Artificial Intelligence Laboratory 100191 China

Scene Graph Generation (SGG) plays a pivotal role in downstream vision-language tasks. Existing SGG methods typically suffer from poor compositional generalizations on unseen triplets. They are generally trained on incompletely annotated scene graphs that contain dominant triplets and tend to bias toward these seen triplets during inference. To address this issue, we propose a Triplet Calibration and Reduction (T-CAR) framework in this paper. In our framework, a triplet calibration loss is first presented to regularize the representations of diverse triplets and to simultaneously excavate the unseen triplets in incompletely annotated training scene graphs. Moreover, the unseen space of scene graphs is usually several times larger than the seen space since it contains a huge number of unrealistic compositions. Thus, we propose an unseen space reduction loss to shift the attention of excavation to reasonable unseen compositions to facilitate the model training. Finally, we propose a contextual encoder to improve the compositional generalizations of unseen triplets by explicitly modeling the relative spatial relations between subjects and objects. Extensive experiments show that our approach achieves consistent improvements for zero-shot SGG over state-of-the-art methods. The code is availab.e at https://***/jkli1998/T-CAR. Copyright © 2023, The Authors. All rights reserved.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

AGDB: A Dictionary-based Malicious Domain Detection Method based on Representation Fusion 7

AGDB: A Dictionary-based Malicious Domain Detection Method b...

引用

7th IEEE International Conference on Data science in Cyberspace, DSC 2022

作者： Zhai, You Dong, Hao Li, Zhoujun Yang, Liqun He, Longtao Beihang University State Key Lab of Software Development Environment China Beihang University School of Cyber Science and Technology China National Computer Network Emergency Response Technical Team Coordination Center of China China

ISBN: (数字)9781665474801

ISBN: (纸本)9781665474801

Botnets are starting to use domain generation algorithms (DGAs) extensively to enhance the stealth of command and control (C&C) communications between C&C servers and bots. Domains generated by DGAs are called algorithmically generated domains (AGDs), which also known as malicious domains. Detection of AGDs is a crucial element for fighting botnets and security researchers have proposed a variety of DGA detection methods. In order to avoid the detection of DGA detectors, various types of DGAs are continuously updated. Among them, the dictionary-based malicious domain, with strong camouflage, is the most advanced DGA representative and the previous detection methods are very ineffective on this type of malicious domain. To solve this problem, we explore the dictionary-based malicious domain generation algorithm, and propose, AGDB, a dictionary-based malicious domain detection method based on representation fusion, which combines features extracted from the context-based malicious domain detection model with features extracted from the graph-based malicious domain detection model. The experimental results show that the detection method based on representation fusion significantly outperforms the existing methods in terms of precision and recall. © 2022 IEEE.

关键词： Botnet

来源：评论

学校读者我要写书评

暂无评论

Enhancing the Transferability of Adversarial Attacks with Stealth Preservation

Enhancing the Transferability of Adversarial Attacks with St...

引用

IEEE computer Society Conference on computer Vision and Pattern Recognition Workshops (CVPRW)

作者： Xinwei Zhang Tianyuan Zhang Yitong Zhang Shuangcheng Liu School of Computer Science and Engineering Beihang University Beijing China State Key Lab of Software Development Environment Beihang University Beijing China Shen Yuan Honors College Beihang University Beijing China

ISBN: (数字)9798350365474

ISBN: (纸本)9798350365481

Deep neural networks are susceptible to attacks from adversarial examples in recent years. Especially, the black-box attacks cause a more serious threat to practical applications. However, while most existing black-box attacks have achieved a high success rate in deceiving models, they have not focused on the stealthiness of adversarial examples, often exhibiting suspicious visual appearances. To address this issue, this paper proposes the Mask Momentum Iterative Attack (MMIA), which introduces a masking mechanism and adopts an optimal perturbation strategy to identify regions of an image most vulnerable to attacks. This approach effectively ensures the transferability and stealthiness of adversarial examples. Simultaneously, by integrating image enhancement techniques and temporal and spatial momentum terms into the iterative process of the attack, we prevent the attack from getting stuck in local optima, further improving the transferability of adversarial examples. To enhance the success rate of black-box attacks, we apply MMIA to a model ensemble using a joint optimization strategy. We demonstrate that adversarially trained models with a strong defense ability are also susceptible to our black-box attacks. We conduct extensive experiments on classification tasks using common vision models, and our results significantly demonstrate the superiority of our method over state-of-the-art approaches when considering both transferability and stealthiness.

关键词： Visualization Fuses Computational modeling Perturbation methods Closed box Pattern recognition Iterative methods

来源：评论

学校读者我要写书评

暂无评论

How Well Do Large Language Models Serve as End-to-End Secure Code Producers?

arXiv

引用

arXiv 2024年

作者： Gong, Jianian Duan, Nachuan Tao, Ziheng Gong, Zhaohui Yuan, Yuan Huang, Minlie School of Computer Science and Engineering Beihang University Beijing China School of Computer Science and Engineering Beihang University State Key Laboratory of Software Development Environment Zhongguancun Laboratory Beijing100191 China Tsinghua University Zhongguancun Laboratory Beijing China

The rapid advancement of large language models (LLMs) such as GPT-4 has revolutionized the landscape of software engineering, positioning these models at the core of modern development practices. As we anticipate these models to evolve into the primary and trustworthy tools used in software development, ensuring the security of the code they produce becomes paramount. How well can LLMs serve as end-to-end secure code producers? This paper presents a systematic investigation into LLMs’ inherent potential to generate code with fewer vulnerabilities. Specifically, We studied GPT-3.5 and GPT-4’s capability to identify and repair vulnerabilities in the code generated by four popular LLMs including themselves (GPT-3.5, GPT-4, Code Llama, and CodeGeeX2). By manually or automatically reviewing 4,900 pieces of code, our study reveals that: (1) large language models lack awareness of scenario-relevant security risks, which leads to the generation of over 75% vulnerable code on the SecurityEval benchmark;(2) LLMs such as GPT-3.5 and GPT-4 are unable to precisely identify vulnerabilities in the code they generated;(3) GPT-3.5 and GPT-4 can achieve 33.2%~59.6% success rates in repairing the insecure code produced by the 4 LLMs, but they both perform poorly when repairing self-produced code, indicating self-repair "blind spots". To address the limitation of a single round of repair, we developed a lightweight tool that prompts LLMs to construct safer source code through an iterative repair procedure based on the insights gained from our study. Experiments show that assisted by semantic analysis engines, our tool significantly improves the success rates of repair to 65.9%~85.5%. Copyright © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Generalizing 6-DoF Grasp Detection via Domain Prior Knowledge

Generalizing 6-DoF Grasp Detection via Domain Prior Knowledg...

引用

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Haoxiang Ma Modi Shi Boyang Gao Di Huang State Key Laboratory of Software Development Environment Beihang University Beijing China School of Computer Science and Engineering Beihang University Beijing China School of Computer Science and Technology Harbin Institute of Technology Harbin China Geometry Robotics

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

We focus on the generalization ability of the 6-DoF grasp detection method in this paper. While learning-based grasp detection methods can predict grasp poses for unseen ob-jects using the grasp distribution learned from the training set, they often exhibit a significant performance drop when encountering objects with diverse shapes and struc-tures. To enhance the grasp detection methods' general-ization ability, we incorporate domain prior knowledge of robotic grasping, enabling better adaptation to objects with significant shape and structure differences. More specifi-cally, we employ the physical constraint regularization during the training phase to guide the model towards predicting grasps that comply with the physical rule on grasping. For the unstable grasp poses predicted on novel objects, we design a contact-score joint optimization using the pro-jection contact map to refine these poses in cluttered sce-narios. Extensive experiments conducted on the GraspNet-1 billion benchmark demonstrate a substantial performance gain on the novel object set and the real-world grasping experiments also demonstrate the effectiveness of our gen-eralizing 6-DoF grasp detection method. Code is availab.e at https://***/mahaoxiang822/Generalizing-Grasp.

关键词： Training computer vision Shape Grasping Benchmark testing Predictive models Performance gain

来源：评论

学校读者我要写书评

暂无评论

Non-salient region erasure for time series augmentation

引用

Frontiers of computer science 2022年第6期16卷 203-205页

作者： Pin LIU Xiaohui GUO Bin SHI Rui WANG Tianyu WO Xudong LIU State Key Lab of Software Development Environment(SKLSDE) Beihang UniversityBeijing100191China Beijing Advanced Innovation Center for Big Data and Brain Computing(BDBC) Beijing100191China Hangzhou Innovation Institute Beihang UniversityHangzhou310052China School of Computer Science and Technology Xi’an Jiaotong UniversityXi’an710049China

1 Introduction Time seriesaugmentationis an essential approachto solvethe overfitting problem on the time series classification(TSC)task[1,2].Although existing approaches perform better in mitigating this problem,none of them focus on protecting saliency regions on time *** key informative shapelets contained in these regions are the core basis for distinguishing categories(e.g.,upward spikes in ECG and high amplitude in Sensor).

关键词： Sensor series series

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：