检索结果-内蒙古大学图书馆

IEEE Workshop on Applications of Computer Vision (WACV)

作者： Thomas Walker Octave Mariotti Amir Vaxman Hakan Bilen University Of Edinburgh

ISBN: (数字)9798331510831

ISBN: (纸本)9798331510848

Positional encodings are a common component of neural scene reconstruction methods, and provide a way to bias the learning of neural fields towards coarser or finer representations. Current neural surface reconstruction methods use a “one-size-fits-all” approach to encoding, choosing a fixed set of encoding functions, and therefore bias, across all scenes. Current state-of-the-art surface reconstruction approaches leverage grid-based multi-resolution hash encoding in order to recover high-detail geometry. We propose a learned approach which allows the network to choose its encoding basis as a function of space, by masking the contribution of features stored at separate grid resolutions. The resulting spatially adaptive approach allows the network to fit a wider range of frequencies without introducing noise. We test our approach on standard benchmark surface reconstruction datasets and achieve state-of-the-art performance on two benchmark datasets.

关键词： Geometry Surface reconstruction Computer vision Noise reconstruction algorithms Benchmark testing Encoding Surface fitting Spatial resolution Standards

来源：评论

学校读者我要写书评

暂无评论

Joint reconstruction of Activity and Attenuation for PET Imaging with Diffusion Prior

Joint Reconstruction of Activity and Attenuation for PET Ima...

引用

IEEE International Symposium on Biomedical Imaging

作者： Sangjin Bae Jae Sung Lee Kuang Gong Seoul National University University of Florida Seoul National University Hospital

ISBN: (数字)9798331520526

ISBN: (纸本)9798331520533

Attenuation correction is essential for accurate quantification in PET imaging. The popular Maximum Likelihood reconstruction of Activity and Attenuation (MLAA) algorithm estimates the attenuation maps and PET activity images jointly from PET sinograms but often suffers from slow convergence and high noise levels. In this work, we developed a diffusion model-based joint reconstruction algorithm that incorporates a diffusion model into the joint reconstruction of attenuation and activity, termed DEMAA. DEMAA requires only a dif-fusion model pre-trained on the widely available CT images, eliminating the need for paired datasets or tracer-specific training. Our preliminary results show that DEMAA significantly accelerated convergence in terms of Root Mean Squared Error (RMSE) and demonstrated potential for jointly estimating attenuation map and activity from non-TOF PET sinogram.

关键词： Training Maximum likelihood estimation Neural networks reconstruction algorithms Attenuation Diffusion models Positron emission tomography Image reconstruction Convergence Noise level

来源：评论

学校读者我要写书评

暂无评论

Single-View reconstruction via Decoupled 3D Gaussian Splatting

Single-View Reconstruction via Decoupled 3D Gaussian Splatti...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Sheng Liu Shiming Zhu Huilong Pi Yunchuan Qin Zhuo Tang Ruihui Li College of Computer Science and Electronic Engineering Hunan University Changsha China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Creating high-quality 3D object representations from a single-view image is challenging. Existing methods tend to infer the geometry and texture information simultaneously within a shared network. However, decoding geometry and texture from a unified network often leads to their entanglement, causing geometric structure collapse or floating artifacts. After revisiting this task, we propose a single-view reconstruction framework based on 3D Gaussian Splatting. The key idea is to decouple Gaussian position attribute generation from texture feature generation. Technically, our framework combines a Geometry Generator, a Texture Generator, and a Gaussian Attributes Decoder. Two parallel branches, Geometry Generator and Texture Generator, aim for point cloud prediction and texture optimization, respectively. Then the Gaussian Attributes Decoder integrates the generated position and texture attributes into a coherent Gaussian point cloud, facilitating efficient novel view synthesis. Extensive qualitative and quantitative evaluations of public datasets demonstrate that our method consistently outperforms existing methods in terms of reconstruction quality and inferring efficiency.

关键词： Geometry Point cloud compression Three-dimensional displays Signal processing reconstruction algorithms Generators Decoding Speech processing Image reconstruction Optimization

来源：评论

学校读者我要写书评

暂无评论

When are Diffusion Priors Helpful in Sparse reconstruction? A Study with Sparse-View CT

When are Diffusion Priors Helpful in Sparse Reconstruction? ...

引用

IEEE International Symposium on Biomedical Imaging

作者： Matt Y. Cheung Sophia Zorek Tucker J. Netherton Laurence E. Court Sadeer Al-Kindi Ashok Veeraraghavan Guha Balakrishnan Department of Electrical & Computer Engineering Rice University Houston TX Department of Radiation Physics The University of Texas MD Anderson Cancer Center Houston TX Department of Statistics Rice University Houston TX Department of Cardiology DeBakey Heart and Vascular Center Houston TX

ISBN: (数字)9798331520526

ISBN: (纸本)9798331520533

Diffusion models demonstrate state-of-the-art performance on image generation, and are gaining traction for sparse medical image reconstruction tasks. However, compared to classical reconstruction algorithms relying on simple ana-lytical priors, diffusion models have the dangerous property of producing realistic looking results even when incorrect, particularly with few observations. We investigate the utility of diffusion models as priors for image reconstruction by varying the number of observations and comparing their performance to classical priors (sparse and Tikhonov regularization) using pixel-based, structural, and downstream metrics. We make comparisons on low-dose chest wall computed tomography (CT) for fat mass quantification. First, we find that classical priors are superior to diffusion priors when the number of projections is “sufficient”. Second, we find that diffusion priors can capture a large amount of detail with very few observations, significantly outperforming classical priors. However, they fall short of capturing all details, even with many observations. Finally, we find that the performance of diffusion priors plateau after extremely few (≈10–15) projections. Ultimately, our work highlights potential issues with diffusion-based sparse reconstruction and underscores the importance of further investigation, particularly in high-stakes clinical settings.

关键词： Measurement Image synthesis Computed tomography reconstruction algorithms Diffusion models Fats Image reconstruction Biomedical imaging

来源：评论

学校读者我要写书评

暂无评论

Compartment Model-Guided Unsupervised reconstruction for Low-Count Dynamic PET Imaging

Compartment Model-Guided Unsupervised Reconstruction for Low...

引用

IEEE International Symposium on Biomedical Imaging

作者： Hengjia Ran Min Guo Huafeng Liu State Key Laboratory of Extreme Photonics and Instrumentation College of Optical Science and Engineering Zhejiang University Hangzhou China

ISBN: (数字)9798331520526

ISBN: (纸本)9798331520533

Image reconstruction from low-count dynamic Positron Emission Tomography (PET) data is challenging due to the ill-conditioning of the inverse problem and poor statistics. This paper introduces a novel compartment model-guided unsupervised method that captures spatio-temporal correlations between images. By predicting tracer concentrations in each compartment, we generate prior images projected into the sinogram domain for comparison with measured data. We integrate physiological, boundary, and data constraints into the loss function while updating kinetic and network parameters to ensure accurate solution propagation. Our method enhances image quality under low-count conditions and in early frames, effectively combining physiological and physical constraints. This approach improves interpretability and generalizability while eliminating the need for training datasets. Experimental results demonstrate its superior performance compared to other reconstruction techniques.

关键词： Training Image quality Biological system modeling reconstruction algorithms Propagation losses Physiology Kinetic theory Positron emission tomography Image reconstruction Unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Enhance Cardiac MRI reconstruction via Unfolding Network with Hybrid Mamba-LSTM

Enhance Cardiac MRI Reconstruction via Unfolding Network wit...

引用

IEEE International Symposium on Biomedical Imaging

作者： Bangjun Li Wenmiao Wang Kemeng Tao Yankun Cao Yujun Li Zhi Liu School of Information Science and Engineering Shandong University Qingdao China School of Software and C-FAIR Shandong University Jinan China

ISBN: (数字)9798331520526

ISBN: (纸本)9798331520533

Cardiac Magnetic Resonance Imaging is an important clinical imaging modality for cardiovascular disease diagnosis. However, long scan times and high costs make cardiac imaging inaccessible, especially as it often requires both high spatial and temporal resolution in a limited time. While existing methods have achieved promising performance, they still lack the ability to capture long spatiotemporal relationships. In this paper, we introduce a novel dynamic reconstruction method to accelerate cardiac MRI reconstruction. Our method adopts a deep unrolling structure and introduces a novel spatiotemporal learning model for high fidelity reconstruction. Specifically, inspired by the recently proposed selective structure state space models, which have shown promising performance due to their efficiency and long-range modelling capability, we propose a hybrid Mamba-LSTM (MLSTM) module as the basic cell to fully exploit the spatiotemporal complementary information hidden in scan sequences. Extensive experiments show that the proposed method can reconstruct high quality images from the highly undersampled data and achieves superior performance compared to state-of-the-arts.

关键词： Costs Magnetic resonance imaging Biological system modeling reconstruction algorithms Spatiotemporal phenomena Image restoration Cardiovascular diseases Spatial resolution Image reconstruction Biomedical imaging

来源：评论

学校读者我要写书评

暂无评论

NeuManifold: Neural Watertight Manifold reconstruction with Efficient and High-Quality Rendering Support

NeuManifold: Neural Watertight Manifold Reconstruction with ...

引用

IEEE Workshop on Applications of Computer Vision (WACV)

作者： Xinyue Wei Fanbo Xiang Sai Bi Anpei Chen Kalyan Sunkavalli Zexiang Xu Hao Su University of California San Diego Adobe Research ETH Zürich University of Tübingen

ISBN: (数字)9798331510831

ISBN: (纸本)9798331510848

While existing volumetric rendering approaches provide photorealistic results, extracting high-quality meshes from optimized neural field representations is challenging. Conversely, existing differentiable rasterization-based methods are typically sensitive to initialization and suffer from poor mesh rendering quality. In this paper, we introduce Neu-Manifold, a novel method for reconstructing watertight manifold meshes with high-quality textures from multi-view input images. NeuManifold overcomes the limitations of existing approaches by first learning a neural volumetric field and then refining it through differentiable mesh extraction and surface rendering. To eliminate artifacts and preserve mesh properties during iso-surface extraction, we introduce a novel differentiable marching cubes method. Instead of traditional textures, we use neural textures to enhance rendering quality. To integrate with modern graphics rendering pipelines, we also provide customized GLSL shader support for neural textures. Extensive experiments demonstrate that NeuManifold outperforms existing mesh-based reconstruction methods in both mesh quality and rendering metrics, achieving comparable or superior rendering quality to prior volume-rendering-based methods. The generated results enable real-time, high-quality rendering and seamlessly support numerous graphics pipelines and applications requiring high-quality meshes, such as 3D printing and physical simulation. https://***/neumanifold/.

关键词： Manifolds Graphics Surface reconstruction Pipelines Refining reconstruction algorithms Rendering (computer graphics) Three-dimensional printing Real-time systems Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

algorithms FOR IMAGE-reconstruction FROM PHOTON-LIMITED DATA USING THE TRIPLE CORRELATION

引用

JOURNAL OF THE OPTICAL SOCIETY OF AMERICA A-OPTICS IMAGE SCIENCE AND VISION 1988年第7期5卷 986-992页

作者： NORTHCOTT, MJ AYERS, GR DAINTY, JC Blackett Laboratory Imperial College London SW7 2BZ UK

The triple-correlation technique of speckle imaging has been investigated in the low-light-level regime. Three practical methods for digital triple-correlation processing have been developed. Each of these methods has successfully reconstructed images from computer-simulated speckle interferograms. In this paper we describe and compare the three algorithms.

关键词： Atmospheric turbulence Fourier transforms Large telescopes Low light levels reconstruction algorithms Spatial resolution

来源：评论

学校读者我要写书评

暂无评论

Robust structured light field depth estimation via spinning line operator

Robust structured light field depth estimation via spinning ...

引用

2024 International Conference on Optical and Photonic Engineering, icOPEN 2024

作者： Guo, Wenfeng Wu, Qingyang Zou, Yifeng Shenzhen Technology University 3002 Lantian Road Pingshan District Shenzhen Guangdong China

ISBN: (数字)9781510688117

ISBN: (纸本)9781510688100

Light fields can simultaneously capture the intensity and direction of each light ray, offering enhanced information for depth estimation. Currently, accurately extracting the epipolar plane lines and handling occlusion are the tough problems of the epipolar plane image depth estimation for light field. Passive depth estimation algorithms only relying on texture information from reconstructed object surfaces struggle to reconstruct texture-free areas, and exhibit poor noise resilience and low accuracy. To address these problems, this paper introduces structured light projection to add accurate and stable texture information, thereby improving the measurement accuracy and robustness, and proposes an epipolar plane line extraction algorithm suitable for structured light field data. This paper designs a spinning line operator based on Structured Light Field characteristics, and uses the variance of the points on the epipolar plane line as the cost of measuring the extraction accuracy. It judges and handles the occlusion problem by gradually removing abnormal points to make the variance converge. This paper measures and reconstructs objects such as standard blocks, standard balls, and portraits, and achieves satisfactory reconstruction results. © 2025 SPIE.

关键词： Structured light reconstruction algorithms Interpolation Cameras Image compression Optical spheres Chaos Image analysis Shadows Spatial resolution

来源：评论

学校读者我要写书评

暂无评论

Deformation reconstruction of Plate Structure Based on FBG tilt Sensor

Deformation Reconstruction of Plate Structure Based on FBG t...

引用

Sensors and Information Technology (ICSI), International Conference on

作者： Jianjun Pan Yahong Li Wei Hou Renxin Wu National Engineering Research Center of Fiber Optic Sensing Technology and Networks Wuhan University of Technology Wuhan China School of Mechanical and Electrical Engineering Wuhan University of Technology Wuhan China School of Information Engineering Wuhan University of Technology Wuhan China

ISBN: (数字)9798331542887

ISBN: (纸本)9798331542894

In this paper, a high sensitivity fiber Bragg grating (FBG) tilt sensor with 281.7 pm/° is designed, which is used to monitor plane deformation. To make the reconstruction result more accurate, this paper optimizes the trigonometric function reconstruction algorithm, and calculates the plane deformation from the x direction and the y direction respectively. In the experiment, the force is applied at the loading position of the thin plate to deform it, and the optimized trigonometric function algorithm is used to reconstruct the thin plate. The error of the final reconstruction result is within 5 %. The results show that the use of FBG tilt sensor can accurately monitor the deformation of the plane.

关键词： Sensitivity Deformation Loading Force Metals reconstruction algorithms Logic gates Fiber gratings Information technology Monitoring

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：