检索结果-内蒙古大学图书馆

16th International Conference on Signal Processing systems, ICSPS 2024

作者： Lyu, Wenqi Ke, Wei Sheng, Hao Ma, Xiao Liu, Lixue Faculty of Applied Sciences Macao Polytechnic University China State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing100191 China

ISBN: (纸本)9781510689251

Light field cameras hold significant value in applications such as depth estimation, 3D video acquisition, and image super-resolution. Compared to traditional single-image super-resolution methods, light field images can capture object information from multiple spatial and angular perspectives, which is more conducive to capturing detail information. However, current advanced light field super-resolution models still exhibit issues like blurriness, jaggedness, and dimness in 4x light field super-resolution. To address this, we propose a novel light field super-resolution framework that utilizes the Retinex decomposition concept to solve common problems such as ghosting and overlap, thereby improving image quality. By decomposing the light field image into illumination and reflection components and processing them separately, our model significantly enhances detail preservation and color consistency. Experimental results on multiple datasets demonstrate that our model exhibits outstanding performance, both subjectively and objectively, highlighting the effectiveness of this framework and its broad application potential in the field of computational imaging. © 2025 SPIE.

关键词： Image acquisition

来源：评论

学校读者我要写书评

暂无评论

Detail Enhancement for Free Surface LBM Using Adaptive Sizing of Coupled Particles 1st

Detail Enhancement for Free Surface LBM Using Adaptive Sizi...

引用

1st International Conference on Extended reality, ICXR 2024

作者： Qu, Qingyue Liu, Huiwen Zhu, Shaonan Hao, Aimin Yu, Peng Gao, Yang State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China Hubei Key Laboratory of Intelligent Yangtze and Hydroelectric Science Yangtze Power Co. Ltd. Hubei Yichang443000 China

ISBN: (纸本)9789819636785

Physically based fluid simulation plays an important role in the immersion and realistic of VR and AR applications. However, efficiency is the bottleneck problem that restricts physical applications. Among those physics-related fluid models, the Lattice Boltzmann Method (LBM), with its localized algorithm, offers high efficiency. However, as a purely Eulerian grid-based approach, achieving rich detail necessitates high resolution. Conversely, Lagrange-based methods can capture intricate details but require substantial computational time to solve large-scale linear systems and manage numerous particle operations. To address these challenges, we propose a hybrid solver that couples the efficient LBM solver with an adaptive particle interpolation method for detail enhancement. Our innovative adaptive particle strategy further reduces particle count while preserving surface detail and improving temporal efficiency by dynamically adjusting particle size based on particle position, distance, and velocity. Experiments demonstrate that our method can display more surface details than LBM alone, delivering visual effects comparable to direct particle interpolation but with higher efficiency. Moreover, our method yields superior visual results compared to adaptive particle methods based on the FLIP approach. This approach is particularly well-suited for large-scale simulations, enhancing both visual quality and time efficiency, and shows great potential for real-time fluid simulations in VR applications. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Computational fluid dynamics

来源：评论

学校读者我要写书评

暂无评论

SIE-DepthNet: Semantic-Guided Monocular Depth Estimation for Dynamic Environment 1st

SIE-DepthNet: Semantic-Guided Monocular Depth Estimation for...

引用

1st International Conference on Extended reality, ICXR 2024

作者： Song, Zilong Gao, Yang Dai, Sijia Li, Shuai Hao, Aimin Zhang, Shoulong State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing China China Nuclear Power Engineering Co. Ltd. Hebei Branch Ningde China Zhongguancun Laboratory Beijing China

ISBN: (纸本)9789819636785

Depth estimation is a fundamental aspect of spatial computation in virtual-real fusion contexts, wherein self-supervised monocular depth estimation presents a persistent challenge within the fields of computer vision and augmented/mixed reality (AR/XR) applications. However, existing methods face difficulties with moving objects, occlusions, and motion blur, resulting in inaccurate estimates, particularly in dynamic regions. Previous solutions either omit challenging areas in the training process or employ pseudo-depth labels, which still cannot fully address the issues. In this paper, we introduce our semantically implicit and explicit depth estimation network, SIE-DepthNet, a novel approach that leverages semantic information to improve monocular depth estimation in dynamic scenes. We articulate a Depth Semantic Feature Fusion Network (DSFFNet) to provide implicit guidance, ensuring consistent depth distributions across categories. Additionally, we propose a semantic-guided ranking loss that optimizes depth accuracy by aligning estimated values with segmentation boundaries, using uncertainty-aware weighting to mitigate the impact of segmentation noise. Extensive experimental results demonstrate that SIE-DepthNet achieves accurate depth map predictions, significantly enhancing performance in dynamic and static scenarios. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Attention mechanism Monocular depth estimation Semantic-guided self-supervised depth estimation

来源：评论

学校读者我要写书评

暂无评论

JGHand: Joint-Driven Animatable Hand Avater via 3D Gaussian Splatting

arXiv

引用

arXiv 2025年

作者： Sun, Zhoutao Shen, Xukun Hu, Yong Zhong, Yuyou Zhou, Xueyang State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China

Since hands are the primary interface in daily interactions, modeling high-quality digital human hands and rendering realistic images is a critical research problem. Furthermore, considering the requirements of interactive and rendering applications, it is essential to achieve real-time rendering and driveability of the digital model without compromising rendering quality. Thus, we propose Jointly 3D Gaussian Hand (JGHand), a novel joint-driven 3D Gaussian Splatting (3DGS)-based hand representation that renders high-fidelity hand images in real-time for various poses and characters. Distinct from existing articulated neural rendering techniques, we introduce a differentiable process for spatial transformations based on 3D key points. This process supports deformations from the canonical template to a mesh with arbitrary bone lengths and poses. Additionally, we propose a real-time shadow simulation method based on per-pixel depth to simulate self-occlusion shadows caused by finger movements. Finally, we embed the hand prior and propose an animatable 3DGS representation of the hand driven solely by 3D key points. We validate the effectiveness of each component of our approach through comprehensive ablation studies. Experimental results on public datasets demonstrate that JGHand achieves real-time rendering speeds with enhanced quality, surpassing state-of-the-art methods. Copyright © 2025, The Authors. All rights reserved.

关键词： Digital elevation model

来源：评论

学校读者我要写书评

暂无评论

Vision-Language Model for Object Detection and Segmentation: A Review and Evaluation

arXiv

引用

arXiv 2025年

作者： Feng, Yongchao Liu, Yajie Yang, Shuai Cai, Wenrui Zhang, Jinqing Zhan, Qiqi Huang, Ziyue Yan, Hongxi Wan, Qiao Liu, Chenguang Wang, Junzhe Lv, Jiahui Liu, Ziqi Shi, Tengyuan Liu, Qingjie Wang, Yunhong State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China

Vision-Language Model (VLM) have gained widespread adoption in Open-Vocabulary (OV) object detection and segmentation tasks. Despite they have shown promise on OV-related tasks, their effectiveness in conventional vision tasks has thus far been unevaluated. In this work, we present the systematic review of VLM-based detection and segmentation, view VLM as the foundational model and conduct comprehensive evaluations across multiple downstream tasks for the first time: 1) The evaluation spans eight detection scenarios (closed-set detection, domain adaptation, crowded objects, etc.) and eight segmentation scenarios (few-shot, open-world, small object, etc.), revealing distinct performance advantages and limitations of various VLM architectures across tasks. 2) As for detection tasks, we evaluate VLMs under three finetuning granularities: zero prediction, visual fine-tuning, and text prompt, and further analyze how different finetuning strategies impact performance under varied task. 3) Based on empirical findings, we provide in-depth analysis of the correlations between task characteristics, model architectures, and training methodologies, offering insights for future VLM design. 4) We believe that this work shall be valuable to the pattern recognition experts working in the fields of computer vision, multimodal learning, and vision foundation models by introducing them to the problem, and familiarizing them with the current status of the progress while providing promising directions for future research. A project associated with this review and evaluation has been created at https://***/better-chao/perceptual_abilities_evaluation. © 2025, CC BY.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

LipText: Lip Tracking Based Text Entry in VR 1st

LipText: Lip Tracking Based Text Entry in VR

引用

1st International Conference on Extended reality, ICXR 2024

作者： Leng, Jiaye Wang, Zijun Wu, Jian Wang, Lili School of Creative Media City University of Hong Kong Kowloon Tong Hong Kong State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing100191 China

ISBN: (纸本)9789819636785

Text entry is an important task in virtual reality (VR), and most existing methods require hand involvement, while hands-free typing has great potential for applications in mobile scenarios. Existing hands-free text entry methods are usually implemented by combining the head and eyes with techniques such as Dwell, Blink and Gesture, which can easily fatigue the user. In this paper, we propose LipText, a lip-tracking-based text entry method in VR. We use a neural network to perform letter-level prediction on the lip data captured by the facial tracker and use head-based selection as an auxiliary to improve the accuracy. We conduct a user study to evaluate our method, the results show a typing speed of 8.63 WPM for the novice group, 9.81 WPM for the potential expert group, and the highest recorded typing speed is 11.13 WPM achieved by a potential expert. Our method is also novice-friendly, and their typing speed increased by 64.38% over a six-day practice. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： virtual environments

来源：评论

学校读者我要写书评

暂无评论

Cluster-Detection-Based Local Super-Resolution Network for Remote Sensing Image Enhancement 1st

Cluster-Detection-Based Local Super-Resolution Network for ...

引用

1st International Conference on Extended reality, ICXR 2024

作者： Li, Lu Zhu, Xia Ni, Shaofeng Gao, Fan Cao, Dinglun Pei, Ziyi Li, Shuai State Grid Location based Service Co. Ltd Beijing China State Grid Information & Telecommunication Group Co. Ltd. Beijing China State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing China

ISBN: (纸本)9789819636785

Remote sensing images are essential for various applications such as environmental monitoring and urban planning. However, the spatial resolution of these images often hinders the accurate detection of small-scale or densely packed targets. In this work, we advocate a novel method that introduces clustering into local super-resolution networks to enhance image resolution and improve the detectability of small-scale and dense regions. First, K-means clustering is applied to identify dense areas based on both the color and spatial properties of image pixels, efficiently pinpointing regions where small objects are concentrated. Next, a lightweight cross-attention-based super-resolution network is employed to enhance the resolution of these key regions, which improves object detection accuracy. The proposed method is computationally efficient, incorporating depthwise separable convolutions and focusing the cross-attention mechanism on regions of interest, thereby minimizing overhead. Extensive experiments on the WV-3 and DOTA datasets demonstrate that our method significantly improves image quality, outperforming existing methods in terms of both PSNR and SSIM. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Optical remote sensing

来源：评论

学校读者我要写书评

暂无评论

Text-driven 3D Human Generation via Contrastive Preference Optimization

arXiv

引用

arXiv 2025年

作者： Zhou, Pengfei Shen, Xukun Hu, Yong State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China

Recent advancements in Score Distillation Sampling (SDS) have significantly improved text-driven 3D human generation, enabling plausible synthesis from natural language descriptions. However, existing methods struggle with fine-grained semantic alignment, leading to inconsistencies in structure, texture, and attribute representation. To address these challenges, we propose a contrastive preference modeling framework that integrates both positive and negative prompts to enhance SDS. This approach refines text-conditioned generation, improving semantic consistency and 3D fidelity. Specifically, we introduce a preference optimization module that aggregates knowledge from multiple preference models to ensure accurate representation of intricate attributes and nuanced details. Additionally, we propose a negation preference module with static-dynamic negation prompts to systematically distinguish relevant details from extraneous elements, ensuring geometric and textural consistency. To further mitigate artifacts and prevent reward hacking, we develop an LLM-driven dynamic negation prompting mechanism that analyzes textual inputs, decomposes complex attributes, and filters irrelevant components. Extensive experiments demonstrate that our method achieves state-of-the-art performance, significantly enhancing texture realism, structure coherence, and overall visual alignment, particularly for complex and lengthy textual descriptions. © 2025, CC BY.

关键词： Three dimensional computer graphics

来源：评论

学校读者我要写书评

暂无评论

Dense object detection for remote sensing images based on multi-scale partitioning and super-resolution optimization 24

Dense object detection for remote sensing images based on mu...

引用

3rd International Conference on Signal Processing, Computer Networks and Communications, SPCNC 2024

作者： Li, Lu Zhu, Xia Ni, Shaofeng Gao, Fan State Grid Location based Service Co. Ltd. Beijing China State Grid Information & Telecommunication Group Co. Ltd. Beijing China State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing China

ISBN: (纸本)9798400710834

Detecting dense objects in remote sensing image is a critical yet challenging task due to variations in target scale, image resolution, and complex backgrounds. Traditional detection methods often struggle with these dense regions, particularly for small or overlapping objects. To overcome these limitations, this study introduces an innovative approach that combines multi-scale partitioning with super-resolution optimization to improve the detection of densely packed objects. By applying K-Means clustering to identify regions of interest (ROI) and using a local super-resolution network to enhance these areas, we improve the clarity and detectability of small-scale objects. Additionally, we implement a global-local feature extraction architecture, where global features are captured using Transformer-based attention, while local features focus on fine-grained object details. Non-Maximum Suppression (NMS) is employed to fuse multi-scale features and eliminate redundant detections. Experimental results on the DOTA and HRSC2016 datasets demonstrate the effectiveness of our approach, achieving superior performance over state-of-the-art methods in detecting small and densely packed targets. This method shows significant promise for remote sensing applications requiring high-precision target detection. © 2024 Copyright held by the owner/author(s).

关键词： Proximity sensors

来源：评论

学校读者我要写书评

暂无评论

Exploring the Overall Stiffness Distribution Law of Weakly Rigid Composite Fairings 7th

Exploring the Overall Stiffness Distribution Law of Weakly R...

引用

7th China Aeronautical Science and technology Conference, CASTC 2024

作者： Yin, Rongying Kong, Chuifeng Dong, Yushuang Ma, Jichuan Gao, Hanjun Harbin Hafei Aviation Industry Co Ltd Harbin150060 China School of Mechanical Engineering and Automation State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China School of Technology Forestry University Beijing100083 China

ISBN: (纸本)9789819797677

To elucidate the stresses and deformations occurring during the assembly of the powerhouse fairing, finite element analysis (ABAQUS) was employed to emulate the assembly process. A numerical simulation model was constructed for the assembly process of the weakly rigid composite fairing by digitizing its geometry, material properties, and the forces exerted during assembly. Initially, the geometric data and material parameters of the fairing were meticulously collected and analyzed, considering its weakly rigid composite material properties. Subsequently, the finite element analysis method was utilized to partition the fairing’s geometry into minute regions, thereby establishing a numerical simulation model of the fairing assembly process. During the simulation, various forces acting on the fairing during assembly, such as the binding force of the assembly fixture and mechanical loads, were taken into account. The stress and deformation distributions during the fairing assembly process were derived through numerical simulation of the model. The findings indicate that the overall stiffness distribution law of the weakly rigid composite fairing can be summarized as follows: larger fixed boundaries, smaller applied forces, more rigid materials, and more intricate geometries enhance the overall stiffness of the fairing. © Chinese Society of Aeronautics and Astronautics 2025.

关键词： ABAQUS

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：