向量-矩阵张量环辐射场新视图合成模型

李莹戈; 龙珍; 苟艺馨; 林薪雨; 朱策

doi:10.12263/DZXB.20260225

您当前的位置：

首页 >

文章列表页 >

向量-矩阵张量环辐射场新视图合成模型

更新时间：2026-06-15

- 向量-矩阵张量环辐射场新视图合成模型
- Vector-Matrix Tensor Ring Radiance Fields for Novel View Synthesis
- 电子学报 2026年页码：1-12
- 作者机构：
  
  1.电子科技大学信息与通信工程学院，四川成都 611731
  2.重庆大学微电子与通信工程学院，重庆 401331
- 作者简介：
  
  李莹戈女，2002年12月出生于河南省鹤壁市。现为电子科技大学信息与通信工程学院硕士研究生。主要研究方向为张量信号处理、三维场景重建。E-mail: 202322011928@std.uestc.edu.cn
  龙珍女，1993年6月出生于四川省内江市。现为电子科技大学信息与通信工程学院副教授、硕士生导师。主要研究方向为张量信号处理、三维场景重建。E-mail: zhen.long@uestc.edu.cn
  苟艺馨女，1999年10月出生于甘肃省庆阳市。现为电子科技大学信息与通信工程学院博士研究生。主要研究方向为高维数据表征、三维场景重建。E-mail: gyx@std.uestc.edu.cn
  林薪雨男，1991年12月出生于四川省德阳市。现为重庆大学微电子与通信工程学院副教授、硕士生导师。主要研究方向为视觉高精度定位、计算机视觉与信号处理、具身智能（无人驾驶与工业机器人）。E-mail: xinyulin@cqu.edu.cn
  朱策男，1969年9月出生于四川省自贡市。现为电子科技大学信息与通信工程学院教授、博士生导师。主要研究方向为计算机图像与视频处理。E-mail: eczhu@uestc.edu.cn
- 基金信息：
  
  四川省自然科学基金(2025ZNSFSC0002);国家自然科学基金(62401102;62401112)
- DOI：10.12263/DZXB.20260225
  中图分类号： TP391;
- 收稿：2026-04-13，
  
  录用：2026-05-20，
  
  网络首发：2026-06-15，
- 稿件说明：
移动端阅览
李莹戈, 龙珍, 苟艺馨, 等. 向量-矩阵张量环辐射场新视图合成模型[J/OL]. 电子学报, 2026,1-12.

LI Yingge, LONG Zhen, GOU Yixin, et al. Vector-Matrix Tensor Ring Radiance Fields for Novel View Synthesis[J/OL]. ACTA ELECTRONICA SINICA, 2026, 1-12.
李莹戈, 龙珍, 苟艺馨, 等. 向量-矩阵张量环辐射场新视图合成模型[J/OL]. 电子学报, 2026,1-12. DOI： 10.12263/DZXB.20260225.

LI Yingge, LONG Zhen, GOU Yixin, et al. Vector-Matrix Tensor Ring Radiance Fields for Novel View Synthesis[J/OL]. ACTA ELECTRONICA SINICA, 2026, 1-12. DOI： 10.12263/DZXB.20260225.

摘要

基于张量的辐射场方法通过张量回归建立输入（三维空间位置）与输出（体密度、外观特征）之间的映射关系，依托紧凑的场景表示，在保持高质量渲染效果的同时，显著提升了新视图合成效率。然而，现有方法无论是采用传统张量分解还是张量链（Tensor Train，TT）分解，均难以充分挖掘三维场景空间结构信息，对场景深层特征刻画不足。针对这一问题，本文在向量-矩阵（Vector-Matrix，VM）分解框架的基础上，引入张量环（Tensor Ring，TR）分解，提出了向量-矩阵张量环辐射场（Vector-Matrix Tensor Ring Radiance Fields，VMTR-RF）模型用于新视图合成。与现有的张量辐射场方法不同，VMTR分解采用分层建模策略：首先，利用VM分解将场景表示为一系列向量与矩阵因子外积的组合，实现对三维场景的初步紧凑表示；随后，将向量矩阵因子重组为高阶张量，并利用TR分解将其表示为多个核张量构成的张量环网络，从而更充分地捕获三维场景深层特征信息。得益于VMTR分解的优势，VMTR-RF在体密度估计和外观特征学习方面表现出更强的建模能力；最后，利用体渲染技术，结合学习到的体密度与外观特征合成新视图。实验结果表明，VMTR-RF优于现有最先进方法，尤其在保持细节方面表现突出，能够更好地重建锐利边缘、复杂结构和自然纹理，在保持紧凑场景表示的同时实现了更高质量的新视图合成结果。

Abstract

Tensor-based radiance field methods established a mapping between inputs (3D spatial positions) and outputs (volume density and appearance features) via tensor regression. These methods relied on compact scene representations and significantly improved the efficiency of novel view synthesis while maintaining high-quality rendering results. However

existing approaches

whether based on conventional tensor decomposition or tensor train (TT) decomposition

are unable to fully exploit structural information in 3D scene space

thereby limiting the representation of deep-level features. To address this issue

we introduced tensor ring (TR) decomposition into the vector-matrix (VM) decomposition framework and proposed a vector-matrix tensor ring radiance fields (VMTR-RF) model for novel view synthesis. Unlike existing tensor radiance field methods

VMTR decomposition adopted a hierarchical modeling strategy: VM decomposition was first used to represent the scene as a combination of outer products of multiple vector and matrix factors

enabling an initial compact representation of the 3D scene. The vector-matrix factors were then reorganized into high-order tensors and further decomposed using TR decomposition

resulting in a tensor ring network composed of multiple core tensors

thereby enabling more effective capture of deep-level features in 3D scenes. Benefiting from the VMTR decomposition

VMTR-RF exhibited stronger modeling capability in volume density estimation and appearance feature learning. Finally

novel view synthesis was performed using volume rendering by combining the learned volume density and appearance features. Experimental results demonstrated that VMTR-RF outperformed existing state-of-the-art methods

particularly in detail preservation

enabling better reconstruction of sharp edges

complex structures

and natural textures

while achieving higher-quality novel view synthesis with a compact scene representation.

关键词

Keywords

references

Riegler G , Koltun V . Free view synthesis [C ] // Computer Vision - ECCV 2020 . Cham : Springer , 2020 : 623 - 640 . DOI: 10.1007/978-3-030-58529-7_37 http://dx.doi.org/10.1007/978-3-030-58529-7_37

Cai Jintong , Lu Huimin . NeRF-based multi-view synthesis techniques: A survey [C ] // 2024 International Wireless Communications and Mobile Computing . Piscataway : IEEE , 2024 : 208 - 213 . DOI: 10.1109/iwcmc61514.2024.10592441 http://dx.doi.org/10.1109/iwcmc61514.2024.10592441

Gao Chen , Saraf A , Kopf J , et al . Dynamic view synthesis from dynamic monocular video [C ] // 2021 IEEE/CVF International Conference on Computer Vision . Piscataway : IEEE , 2021 : 5692 - 5701 . DOI: 10.1109/iccv48922.2021.00566 http://dx.doi.org/10.1109/iccv48922.2021.00566

Groueix T , Fisher M , Kim V G , et al . A papier-mache approach to learning 3D surface generation [C ] // 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2018 : 216 - 224 . DOI: 10.1109/cvpr.2018.00030 http://dx.doi.org/10.1109/cvpr.2018.00030

Wang Nanyang , Zhang Yinda , Li Zhuwen , et al . Pixel2Mesh: Generating 3D mesh models from single RGB images [C ] // Computer Vision-ECCV 2018 . Cham : Springer International Publishing , 2018 : 55 - 71 . DOI: 10.1007/978-3-030-01252-6_4 http://dx.doi.org/10.1007/978-3-030-01252-6_4

Charles R Q , Su Hao , Mo Kaichun , et al . PointNet: Deep learning on point sets for 3D classification and segmentation [C ] // 2017 IEEE Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2017 : 77 - 85 . DOI: 10.1109/cvpr.2017.16 http://dx.doi.org/10.1109/cvpr.2017.16

Ji Mengqi , Gall J , Zheng Haitian , et al . SurfaceNet: An end-to-end 3D neural network for multiview stereopsis [C ] // 2017 IEEE International Conference on Computer Vision . Piscataway : IEEE , 2017 : 2326 - 2334 . DOI: 10.1109/iccv.2017.253 http://dx.doi.org/10.1109/iccv.2017.253

Qi C R , Su Hao , Nießner M , et al . Volumetric and multi-view CNNs for object classification on 3D data [C ] // 2016 IEEE Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2016 : 5648 - 5656 . DOI: 10.1109/cvpr.2016.609 http://dx.doi.org/10.1109/cvpr.2016.609

Mildenhall B , Srinivasan P P , Tancik M , et al . NeRF: Representing scenes as neural radiance fields for view synthesis [J ] . Communications of the ACM , 2021 , 65 ( 1 ): 99 - 106 . DOI: 10.1145/3503250 http://dx.doi.org/10.1145/3503250

Chan E R , Monteiro M , Kellnhofer P , et al . Pi-GAN: Periodic implicit generative adversarial networks for 3D-aware image synthesis [C ] // 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2021 : 5795 - 5805 . DOI: 10.1109/cvpr46437.2021.00574 http://dx.doi.org/10.1109/cvpr46437.2021.00574

Martin-Brualla R , Radwan N , Sajjadi M S M , et al . NeRF in the wild: Neural radiance fields for unconstrained photo collections [C ] // 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2021 : 7206 - 7215 . DOI: 10.1109/cvpr46437.2021.00713 http://dx.doi.org/10.1109/cvpr46437.2021.00713

Schwarz K , Liao Yiyi , Niemeyer M , et al . GRAF: Generative radiance fields for 3D-aware image synthesis [PP/OL ] . V4. arXiv ( 2021-03-30 )[ 2026-04-10 ] . https://doi.org/10.48550/arXiv.2007.02442 https://doi.org/10.48550/arXiv.2007.02442 .

Xiang Fanbo , Xu Zexiang , Hasan M , et al . NeuTex: Neural texture mapping for volumetric neural rendering [C ] // 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2021 : 7115 - 7124 . DOI: 10.1109/cvpr46437.2021.00704 http://dx.doi.org/10.1109/cvpr46437.2021.00704

Pumarola A , Corona E , Pons-Moll G , et al . D-NeRF: Neural radiance fields for dynamic scenes [C ] // 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2021 : 10313 - 10322 . DOI: 10.1109/cvpr46437.2021.01018 http://dx.doi.org/10.1109/cvpr46437.2021.01018

Adamkiewicz M , Chen T , Caccavale A , et al . Vision-only robot navigation in a neural radiance world [J ] . IEEE Robotics and Automation Letters , 2022 , 7 ( 2 ): 4606 - 4613 . DOI: 10.1109/lra.2022.3150497 http://dx.doi.org/10.1109/lra.2022.3150497

Gordon C , Chng S F , MacDonald L , et al . On quantizing implicit neural representations [C ] // 2023 IEEE/CVF Winter Conference on Applications of Computer Vision . Piscataway : IEEE , 2023 : 341 - 350 . DOI: 10.1109/wacv56688.2023.00042 http://dx.doi.org/10.1109/wacv56688.2023.00042

Zhong Hongliang , Zhang Jingbo , Liao Jing . VQ-NeRF: Neural reflectance decomposition and editing with vector quantization [J ] . IEEE Transactions on Visualization and Computer Graphics , 2024 , 30 ( 9 ): 6247 - 6260 . DOI: 10.1109/tvcg.2023.3330518 http://dx.doi.org/10.1109/tvcg.2023.3330518

Liu Lingjie , Gu Jiatao , Lin K Z , et al . Neural sparse voxel fields [PP/OL ] . V2. arXiv ( 2021-01-06 )[ 2026-04-10 ] . https://doi.org/10.48550/arXiv.2007.11571 https://doi.org/10.48550/arXiv.2007.11571 .

Yu A , Li Ruilong , Tancik M , et al . PlenOctrees for real-time rendering of neural radiance fields [C ] // 2021 IEEE/CVF International Conference on Computer Vision . Piscataway : IEEE , 2021 : 5732 - 5741 . DOI: 10.1109/iccv48922.2021.00570 http://dx.doi.org/10.1109/iccv48922.2021.00570

Fridovich-Keil S , Yu A , Tancik M , et al . Plenoxels: Radiance fields without neural networks [C ] // 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2022 : 5491 - 5500 . DOI: 10.1109/cvpr52688.2022.00542 http://dx.doi.org/10.1109/cvpr52688.2022.00542

Müller T , Evans A , Schied C , et al . Instant neural graphics primitives with a multiresolution hash encoding [J ] . ACM Transactions on Graphics , 2022 , 41 ( 4 ): 1 - 15 . DOI: 10.1145/3528223.3530127 http://dx.doi.org/10.1145/3528223.3530127

Chen Anpei , Xu Zexiang , Geiger A , et al . TensoRF: Tensorial radiance fields [C ] // Computer Vision - ECCV 2022 . Cham : Springer , 2022 : 333 - 350 . DOI: 10.1007/978-3-031-19824-3_20 http://dx.doi.org/10.1007/978-3-031-19824-3_20

Gao Quankai , Xu Qiangeng , Su Hao , et al . Strivec: Sparse tri-vector radiance fields [C ] // 2023 IEEE/CVF International Conference on Computer Vision . Piscataway : IEEE , 2023 : 17523 - 17533 . DOI: 10.1109/iccv51070.2023.01611 http://dx.doi.org/10.1109/iccv51070.2023.01611

Kim S B , Kim S , Ahn D , et al . BTD-RF: 3D scene reconstruction using block-term tensor decomposition [J ] . Applied Intelligence , 2024 , 54 ( 8 ): 6319 - 6332 . DOI: 10.1007/s10489-024-05476-0 http://dx.doi.org/10.1007/s10489-024-05476-0

Cichocki A , Lee N , Oseledets I , et al . Tensor networks for dimensionality reduction and large-scale optimization: Part 1 low-rank tensor decompositions [J ] . Foundations and Trends® in Machine Learning , 2016 , 9 ( 4/5 ): 249 - 429 . DOI: 10.1561/2200000059 http://dx.doi.org/10.1561/2200000059

Cichocki A . Era of big data processing: A new approach via tensor networks and tensor decompositions [PP/OL ] . V4. arXiv ( 2014-08-24 )[ 2026-04-10 ] . https://doi.org/10.48550/arXiv.1403.2048 https://doi.org/10.48550/arXiv.1403.2048 .

Bengua J A , Phien H N , Tuan H D , et al . Efficient tensor completion for color image and video recovery: Low-rank tensor train [J ] . IEEE Transactions on Image Processing , 2017 , 26 ( 5 ): 2466 - 2479 . DOI: 10.1109/tip.2017.2672439 http://dx.doi.org/10.1109/tip.2017.2672439

Obukhov A , Usvyatsov M , Sakaridis C , et al . TT-NF: Tensor train neural fields [J ] . IEEE Journal of Selected Topics in Signal Processing , 2024 , 18 ( 6 ): 1024 - 1035 . DOI: 10.1109/jstsp.2024.3454980 http://dx.doi.org/10.1109/jstsp.2024.3454980

Loeschcke S , Wang Dan , Leth-Espensen C , et al . Coarse-to-fine tensor trains for compact visual representations [PP/OL ] . V1. arXiv ( 2024-06-06 )[ 2026-04-10 ] . https://doi.org/10.48550/arXiv.2406.04332 https://doi.org/10.48550/arXiv.2406.04332 .

Shi Jinglei , Guillemot C . Light field compression via compact neural scene representation [C ] // ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing . Piscataway : IEEE , 2023 : 1 - 5 . DOI: 10.1109/icassp49357.2023.10095668 http://dx.doi.org/10.1109/icassp49357.2023.10095668

Boyko A I , Matrosov M P , Oseledets I V , et al . TT-TSDF: Memory-efficient TSDF with low-rank tensor train decomposition [C ] // 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems . Piscataway : IEEE , 2020 : 10116 - 10121 . DOI: 10.1109/iros45743.2020.9341464 http://dx.doi.org/10.1109/iros45743.2020.9341464

Zhao Qibin , Zhou Guoxu , Xie Shengli , et al . Tensor ring decomposition [PP/OL ] . V1. arXiv ( 2016-06-17 )[ 2026-04-10 ] . https://doi.org/10.48550/arXiv.1606.05535 https://doi.org/10.48550/arXiv.1606.05535 .

Long Zhen , Zhu Ce , Liu Jiani , et al . Bayesian low rank tensor ring for image recovery [J ] . IEEE Transactions on Image Processing , 2021 , 30 : 3568 - 3580 . DOI: 10.1109/tip.2021.3062195 http://dx.doi.org/10.1109/tip.2021.3062195

Liu Sheng , Zhao Xile , Zhang Hao . Block tensor ring decomposition: Theory and application [J ] . IEEE Transactions on Signal Processing , 2025 , 73 : 3029 - 3043 . DOI: 10.1109/tsp.2025.3589059 http://dx.doi.org/10.1109/tsp.2025.3589059

Matsui Y , Yokota T . Broadcast product: Shape-aligned element-wise multiplication and beyond [PP/OL ] . V1. arXiv ( 2024-09-26 )[ 2026-04-10 ] . https://doi.org/10.48550/arXiv.2409.17502 https://doi.org/10.48550/arXiv.2409.17502 .

Knapitsch A , Park J , Zhou Qianyi , et al . Tanks and temples: Benchmarking large-scale scene reconstruction [J ] . ACM Transactions on Graphics , 2017 , 36 ( 4 ): 1 - 13 . DOI: 10.1145/3072959.3073599 http://dx.doi.org/10.1145/3072959.3073599

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于大语言模型语义增强的多模态智能合约漏洞检测方法研究

基于自适应群体掩码图卷积网络的行人轨迹预测

面向类别不平衡ECG的快速患者间域适应心律失常识别方法

非合作对抗场景下的隐真示假调制识别方法

基于国密SM9的分层标识签名方案