检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Wang, Guotao Chen, Chenglizhao Hao, Aimin Qin, Hong Fan, Deng-Ping The State Key Laboratory of Virtual Reality Technology and Systems Beihang University China The College of Computer Science and Technology China University of Petroleum China The State Key Laboratory of Virtual Reality Technology and Systems Beihang University Research Unit of Virtual Human and Virtual Surgery Chinese Academy of Medical Sciences and with Pengcheng Laboratory China The Computer Science Department Stony Brook University United States ETH Zurich Zurich Switzerland

To date, the widely adopted way to perform fixation collection in panoptic video is based on a head-mounted display (HMD), where users' fixations are collected while wearing an HMD to explore the given panoptic scene freely. However, this widely-used data collection method is insufficient for training deep models to accurately predict which regions in a given panoptic are most important when it contains intermittent salient events. The main reason is that there always exist "blind zooms" when using HMD to collect fixations since the users cannot keep spinning their heads to explore the entire panoptic scene all the time. Consequently, the collected fixations tend to be trapped in some local views, leaving the remaining areas to be the "blind zooms". Therefore, fixation data collected using HMD-based methods that accumulate local views cannot accurately represent the overall global importance - the main purpose of fixations - of complex panoptic scenes. To conquer, this paper introduces the auxiliary window with a dynamic blurring (WinDB) fixation collection approach for panoptic video, which doesn't need HMD and is able to well reflect the regional-wise importance degree. Using our WinDB approach, we have released a new PanopticVideo-300 dataset, containing 300 panoptic clips covering over 225 categories. Specifically, since using WinDB to collect fixations is blind zoom free, there exists frequent and intensive "fixation shifting" - a very special phenomenon that has long been overlooked by the previous research - in our new set. Thus, we present an effective fixation shifting network (FishNet) to conquer it. All these new fixation collection tool, dataset, and network could be very potential to open a new age for fixation-related research and applications in 360o environments. © 2023, CC BY.

关键词： Helmet mounted displays

来源：评论

学校读者我要写书评

暂无评论

SVDFed: Enabling Communication-Efficient Federated Learning via Singular-Value-Decomposition 42

SVDFed: Enabling Communication-Efficient Federated Learning ...

引用

42nd IEEE International Conference on computer Communications, INFOCOM 2023

作者： Wang, Haolin Liu, Xuefeng Niu, Jianwei Tang, Shaojie Beihang University State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beijing China Zhongguancun Laboratory Beijing China Zhengzhou University Zhengzhou University Research Institute of Industrial Technology School of Information Engineering Zhengzhou China Naveen Jindal School Management University of Texas at Dallas Richardson United States

ISBN: (纸本)9798350334142

Federated learning (FL) is an emerging paradigm of distributed machine learning. However, when applied to wireless network scenarios, FL usually suffers from high communication cost because clients need to transmit their updated gradients to a server in every training round. Although many gradient compression techniques like sparsification and quantization are proposed, they compress clients' gradients independently, without considering the correlations among gradients. In this paper, we propose SVDFed, a collaborative gradient compression framework for FL. SVDFed utilizes Singular Value Decomposition (SVD) to find a few basis vectors, whose linear combination can well represent clients' gradients at a certain round. Due to the correlations among gradients, these basis vectors can still well approximate new gradients in many subsequent rounds. With the help of basis vectors, clients only need to upload the coefficients of the linear combination to the server, which greatly reduces communication cost. In addition, SVDFed leverages the classical PID (Proportional, Integral, Derivative) control to determine the proper time to update basis vectors to maintain their representation ability. Through experiments, we demonstrate that SVDFed outperforms existing gradient compression methods in FL. For example, compared to a popular gradient quantization method QSGD, SVDFed can reduce the communication overhead by 66 % and pending time by 99 %. © 2023 IEEE.

关键词： Singular value decomposition

来源：评论

学校读者我要写书评

暂无评论

MLLMReID: Multimodal Large Language Model-based Person Re-identification

arXiv

引用

arXiv 2024年

作者： Yang, Shan Zhang, Yongfei Beijing Key Laboratory of Digital Media School of Computer Science and Engineering Beihang University Beijing100191 China State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China

Multimodal large language models (MLLM) have achieved satisfactory results in many tasks. However, their performance in the task of ReID (ReID) has not been explored to date. This paper will investigate how to adapt them for the task of ReID. An intuitive idea is to fine-tune MLLM with ReID image-text datasets, and then use their visual encoder as a backbone for ReID. However, there still exist two apparent issues: (1) Designing instructions for ReID, MLLMs may overfit specific instructions, and designing a variety of instructions will lead to higher costs. (2) When fine-tuning the visual encoder of a MLLM, it is not trained synchronously with the ReID task. As a result, the effectiveness of the visual encoder fine-tuning cannot be directly reflected in the performance of the ReID task. To address these problems, this paper proposes MLLMReID: Multimodal Large Language Model-based ReID. Firstly, we proposed Common Instruction, a simple approach that leverages the essence ability of LLMs to continue writing, avoiding complex and diverse instruction design. Secondly, we propose a multi-task learning-based synchronization module to ensure that the visual encoder of the MLLM is trained synchronously with the ReID task. The experimental results demonstrate the superiority of our method. © 2024, CC BY-NC-ND.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Locomotion perception and redirection

引用

virtual Reality & Intelligent Hardware 2021年第6期3卷 I0003-I0004页

作者： Miao WANG Songhai ZHANG Shimin HU State Key Laboratory of Virtual Reality Technology and Systems Beihang UniversityBeijing 100191China Department of Computer Science and Technology Tsinghua UniversityBeijing 100084China

Locomotion is a fundamental interaction technique that allows free navigation in virtual scenes.A large body of literature has demonstrated that natural locomotion experience can significantly improve the sense of presence and help users achieve better performance in tasks such as searching and *** the upcoming metaverse era,locomotion will play a critical role in the success of large-scale interactive *** and redirection are two tightly connected problems in virtual *** a limited physical condition and a target virtual scene,how to make the perceived locomotion as natural as possible,and further unknowingly redirect the user in the physical world is an open *** this special issue,we have selected five papers that provide their latest updates for research problems of view control,virtual jumping,climbing,and searching.

关键词： redirect interactive motion

来源：评论

学校读者我要写书评

暂无评论

FaceCom: Towards High-fidelity 3D Facial Shape Completion via Optimization and Inpainting Guidance

FaceCom: Towards High-fidelity 3D Facial Shape Completion vi...

引用

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Yinglong Li Hongyu Wu Xiaogang Wang Qingzhao Qin Yijiao Zhao Yong Wang Aimin Hao State Key Laboratory of Virtual Reality Technology and Systems Beihang University College of Computer and Information Science Southwest University Peking University School and Hospital of Stomatology

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

We propose FaceCom, a method for 3D facial shape completion, which delivers high-fidelity results for incomplete facial inputs of arbitrary forms. Unlike end-to-end shape completion methods based on point clouds or voxels, our approach relies on a mesh-based generative network that is easy to optimize, enabling it to handle shape completion for irregular facial scans. We first train a shape generator on a mixed 3D facial dataset containing 2405 identities. Based on the incomplete facial input, we fit complete faces using an optimization approach under image inpainting guidance. The completion results are refined through a post-processing step. FaceCom demonstrates the ability to effectively and naturally complete facial scan data with varying missing regions and degrees of missing areas. Our method can be used in medical prosthetic fabrication and the registration of deficient scanning data. Our experimental results demonstrate that FaceCom achieves exceptional performance in fitting and shape completion tasks. The code is available at https://***/dragonylee/***.

关键词： Training Point cloud compression Three-dimensional displays Shape Fitting Refining Generators

来源：评论

学校读者我要写书评

暂无评论

How to Route CUBIC and BBR Packets in Space

How to Route CUBIC and BBR Packets in Space

引用

International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt)

作者： Shuo Huang Zhiyuan Wang Wenhao Lu Kai Shen Jiayi Zhang Shan Zhang Hongbin Luo School of Computer Science and Engineering Beihang University Beijing China Zhongguancun Laboratory Beijing China State Key Laboratory of Virtual Reality Technology and Systems Beijing China State Key Laboratory of Software Development Environment Beijing China

ISBN: (数字)9783903176652

ISBN: (纸本)9798331508722

Low-earth-orbit satellite constellation (e.g., Starlink) is becoming the indispensable component of future Internet. Due to the mobility nature, ground-satellite links (GSLs) and inter-satellite links (ISLs) are not as stable as the fiber links of wired Internet. Consequently, end-to-end TCP connections traversing the LEO constellation will suffer from packet loss and latency variation caused by GSL handover and ISL failure. However, typical congestion control algorithms (e.g., loss-based CUBIC and bandwidth-estimated BBR) rely on these feedbacks to detect congestion in wired Internet. This leads to performance degradation for CUBIC/BBR in LEO satellite networks. In this paper, we propose a Default-and-Backup Routing (DB-R) mechanism to enhance the performance of CUBIC/BBR in LEO satellite networks. Specifically, DB-R leverages default routing to address the packet loss caused by frequent GSL handover. Moreover, DB- R utilizes backup routing to eliminate the packet loss caused by occasional ISL failure. Overall, DB-R only incurs negligible control overhead, but can effectively improve the performance of CUBIC and BBR in LEO satellite networks. We implement DB-R in Linux kernel and evaluate its performance based on OpenSN (i.e., an open-source libarary for emulating LEO satellite networks). Results show that our DB-R mechanism can improve the average throughput of CUBIC (and BBR, respectively) by up to 37 % (and 18 %, respectively) compared with classic OSPF.

关键词： Satellites Linux Wireless networks Low earth orbit satellites Packet loss Handover Routing Throughput Internet Kernel

来源：评论

学校读者我要写书评

暂无评论

Effects of virtual environment and self-representations on perception and physical performance in redirected jumping

引用

virtual Reality & Intelligent Hardware 2021年第6期3卷 451-469页

作者： Yijun LI Miao WANG Derong JIN Frank STEINICKE Qinping ZHAO State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and EngineeringBeihang UniversityBeijing 100083China Peng Cheng Laboratory Shenzhen 518055China Universiẗat Hamburg Hamburg 21071Germany

Background Redirected jumping(RDJ)allows users to explore virtual environments(VEs)naturally by scaling a small real-world jump to a larger virtual jump with virtual camera motion manipulation,thereby addressing the problem of limited physical space in VR *** RDJ studies have mainly focused on detection threshold ***,the effect VE or self representation(SR)has on the perception or performance of RDJs remains *** In this paper,we report experiments to measure the perception(detection thresholds for gains,presence,embodiment,intrinsic motivation,and cybersickness)and physical performance(heart rate intensity,preparation time,and actual jumping distance)of redirected forward jumping under six different combinations of VE(low and high visual richness)and SRs(invisible,shoes,and human-like).Results Our results indicated that the detection threshold ranges for horizontal translation gains were significantly smaller in the VE with high rather than low visual *** different SRs were applied,our results did not suggest significant differences in detection thresholds,but it did report longer actual jumping distances in the invisible body case compared with the other two *** the high visual richness VE,the preparation time for jumping with a human-like avatar was significantly longer than that with other ***,some correlations were found between perception and physical performance *** All these findings suggest that both VE and SRs influence users'perception and performance in RDJ and must be considered when designing locomotion techniques.

关键词： virtual reality virtual locomotion Redirected jumping

来源：评论

学校读者我要写书评

暂无评论

Dual Temporal Transformers for Fine-Grained Dangerous Action Recognition

Dual Temporal Transformers for Fine-Grained Dangerous Action...

引用

IEEE International Conference on Image Processing

作者： Wenfeng Song Xingliang Jin Yang Ding Yang Gao Xia Hou Computer School of Beijing Information Science and Technology University State Key Laboratory of Virtual Reality Technology and Systems Beihang University Research Unit of Virtual Human and Virtual Surgery (2019RU004) Chinese Academy of Medical Sciences

Recognizing dangerous actions is a critical task in computer vision, especially for surveillance applications. While existing deep learning methods have been successful in confined environments, they struggle with the anomalous and salient variations of human postures in dangerous actions. Additionally, finer-grained dangerous actions require more discriminative cues, adding to the complexity of the task. To address these challenges, we propose a novel solution that models the intrinsic and invariant properties of dangerous actions at multiple temporal semantic levels. Concretely, we propose a Dual Temporal Transformers (DTT) to capture temporal interactions between distinct key points in the human body aggregation from shallow to deep layers, increasing the perception field from local to global, simultaneously. By doing so, our method avoids overfitting to unrelated or minor clues in videos and achieves a generalized representation of abnormal actions. We evaluate our approach on indoor and outdoor environments and found that DTT outperforms existing methods in terms of efficiency and accuracy. Our code and dataset are pubic available on https://***/AveryJohnsonJJ/***.

关键词：

来源：评论

学校读者我要写书评

暂无评论

RGB-D Salient Object Detection with Ubiquitous Target Awareness

RGB-D Salient Object Detection with Ubiquitous Target Awaren...

引用

作者： Zhao, Yifan Zhao, Jiawei Li, Jia Chen, Xiaowu State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing China

Conventional RGB-D salient object detection methods aim to leverage depth as complementary information to find the salient regions in both modalities. However, the salient object detection results heavily rely on the quality of captured depth data which sometimes are unavailable. In this work, we make the first attempt to solve the RGB-D salient object detection problem with a novel depth-awareness framework. This framework only relies on RGB data in the testing phase, utilizing captured depth data as supervision for representation learning. To construct our framework as well as achieving accurate salient detection results, we propose a Ubiquitous Target Awareness (UTA) network to solve three important challenges in RGB-D SOD task: 1) a depth awareness module to excavate depth information and to mine ambiguous regions via adaptive depth-error weights, 2) a spatial-aware cross-modal interaction and a channel-aware cross-level interaction, exploiting the low-level boundary cues and amplifying high-level salient channels, and 3) a gated multi-scale predictor module to perceive the object saliency in different contextual scales. Besides its high performance, our proposed UTA network is depth-free for inference and runs in real-time with 43 FPS. Experimental evidence demonstrates that our proposed network not only surpasses the state-of-the-art methods on five public RGB-D SOD benchmarks by a large margin, but also verifies its extensibility on five public RGB SOD benchmarks. © 1992-2012 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

LFNAT 2023 Challenge on Light Field Depth Estimation: Methods and Results

LFNAT 2023 Challenge on Light Field Depth Estimation: Method...

引用

2023 IEEE/CVF Conference on computer Vision and Pattern Recognition Workshops, CVPRW 2023

作者： Sheng, Hao Liu, Yebin Yu, Jingyi Wu, Gaochang Xiong, Wei Guo, Longzhao Xie, Yanlin Zhang, Shuo Chang, Song Lin, Youfang Chao, Wentao Wang, Xuechun Wang, Guanghui Duan, Fuqing Wang, Tun Yang, Da Cui, Zhenglong Wang, Sizhe Zhao, Mingyuan Wang, Qiong Chen, Qianyu Liang, Zhengyu Wang, Yingqian Yang, Jungang Yang, Xueting Deng, Junli Cong, Ruixuan Chen, Rongshan State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University China Beihang Hangzhou Innovation Institute Yuhang China Faculty of Applied Sciences Macao Polytechnic University China Tsinghua University China Shanghaitech University China State Key Laboratory of Synthetical Automation for Process Industries Northeastern University China Beijing Meet Yuan Co. Ltd China Beijing Key Laboratory of Traffic Data Analysis and Mining School of Computer and Information Technology Beijing Jiaotong University China Beijing Normal University China Toronto Metropolitan University Canada College of Computer Science and Technology Zhejiang University of Technology China National University of Defense Technology China School of Information and Communication Engineering Communication University of China China

ISBN: (纸本)9798350302493

This paper reviews the 1st LFNAT challenge on light field depth estimation, which aims at predicting disparity information of central view image in a light field (i.e., pixel offset between central view image and adjacent view image). Compared to multi-view stereo matching, light field depth estimation emphasizes efficient utilization of the 2D angular information from multiple regularly varying views. This challenge specifies UrbanLF [20] light field dataset as the sole data source. There are two phases in total: submission phase and final evaluation phase, in which 75 registered participants successfully submit their predicted results in the first phase and 7 eligible teams compete in the second phase. The performance of all submissions is carefully reviewed and shown in this paper as a new standard for the current state-of-the-art in light field depth estimation. Moreover, the implementation details of these methods are also provided to stimulate related advanced research. © 2023 IEEE.

关键词： Stereo image processing

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：