3D 视觉

LL3DA: Visual Interactive Instruction Tuning for Omni-3D Understanding, Reasoning, and Planning

Sijin Chen, Xin Chen, Chi Zhang, Mingsheng Li, Gang Yu, Hao Fei, Hongyuan Zhu, Jiayuan Fan, Tao Chen

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Comments Project Page: https://ll3da.github.io/

Geometry Aware Field-to-field Transformations for 3D Semantic Segmentation

Dominik Hollidt, Clinton Wang, Polina Golland, Marc Pollefeys

专题命中 点云 ：NeRF(abstract)；point cloud(abstract)；分类 cs.CV

Comments 8 pages

Multi-CLIP: Contrastive Vision-Language Pre-training for Question Answering tasks in 3D Scenes

Alexandros Delitzas, Maria Parelli, Nikolas Hars, Georgios Vlassis, Sotirios Anagnostidis, Gregor Bachmann, Thomas Hofmann

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Comments The first two authors contributed equally. arXiv admin note: text overlap with arXiv:2304.06061

Self-Distillation for Unsupervised 3D Domain Adaptation

Adriano Cardace, Riccardo Spezialetti, Pierluigi Zama Ramirez, Samuele Salti, Luigi Di Stefano

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Comments WACV 2023, Project Page: https://cvlab-unibo.github.io/FeatureDistillation/

SEFormer: Structure Embedding Transformer for 3D Object Detection

Xiaoyu Feng, Heming Du, Yueqi Duan, Yongpan Liu, Hehe Fan

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

3D Object Detection with a Self-supervised Lidar Scene Flow Backbone

Ekim Yurtsever, Emeç Erçelik, Mingyu Liu, Zhijie Yang, Hanzhen Zhang, Pınar Topçam, Maximilian Listl, Yılmaz Kaan Çaylı, Alois Knoll

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Multiview Stereo with Cascaded Epipolar RAFT

Zeyu Ma, Zachary Teed, Jia Deng

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

MetaSets: Meta-Learning on Point Sets for Generalizable Representations

Chao Huang, Zhangjie Cao, Yunbo Wang, Jianmin Wang, Mingsheng Long

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Comments 13 pages, CVPR 2021

NPBG++: Accelerating Neural Point-Based Graphics

Ruslan Rakhimov, Andrei-Timotei Ardelean, Victor Lempitsky, Evgeny Burnaev

专题命中 点云 ：point cloud(abstract)；novel view synthesis(abstract)；分类 cs.CV

Comments Accepted to CVPR 2022. The project page: https://rakhimovv.github.io/npbgpp/

simCrossTrans: A Simple Cross-Modality Transfer Learning for Object Detection with ConvNets or Vision Transformers

Xiaoke Shen, Ioannis Stamos

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Mitigating the Hubness Problem for Zero-Shot Learning of 3D Objects

Ali Cheraghian, Shafin Rahman, Dylan Campbell, Lars Petersson

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Comments BMVC 2019

Revealing Scenes by Inverting Structure from Motion Reconstructions

Francesco Pittaluga, Sanjeev J. Koppal, Sing Bing Kang, Sudipta N. Sinha

专题命中 点云 ：3D vision(abstract)；point cloud(abstract)；分类 cs.CV

Comments 10 pages, 8 figures, to be published in IEEE Conference on Computer Vision and Pattern Recognition 2019

RFDT-Channel: RGB-LiDAR-Based RF Digital Twin Scene Construction for 28 GHz Indoor Ray-Tracing Channel Simulation

RFDT-Channel：基于RGB-LiDAR的RF数字孪生场景构建用于28 GHz室内射线追踪信道仿真

Chengyang Yao, Cunhua Pan, Jiaming Zeng, Yuquan Sun, Haoyang Weng, Haojian Wang, Hong Ren, Jiangzhou Wang

专题命中 点云 ：Gaussian Splatting(abstract)；point cloud(abstract)

AI总结 提出RFDT-Channel工作流，利用RGB图像和LiDAR点云自动构建射频可计算几何与电磁材料属性的数字孪生场景，用于28 GHz射线追踪仿真，通过材料绑定将有效路径数从约742降至约52，同时保持主路径幅度不变。

Open-source segmentation and biometry dataset using spectrally-multiplexed whole-eye optical coherence tomography

开源的基于光谱复用全眼光学相干断层扫描的分割和生物特征数据集

Ruobing Qian, Catherine Fromm, Pushkar Anand, Kyle Johnson, Zach Willms, Yimin Ding, Weihan Zhang, Ali Behrooz, Mohamed El-Haddad

专题命中 点云 ：3D reconstruction(abstract)；point cloud(abstract)

AI总结 本文提出了一种新颖的光谱复用全眼光学相干断层扫描系统，结合深度学习进行表面分割和3D畸变校正，实现了高精度的3D重建，同时开源了包含6621个处理后的体积的数据集，用于改善相关领域的数据获取瓶颈。

CM-EVS: Sparse Panoramic RGB-D-Pose Data for Complete Scene Coverage

CM-EVS：稀疏全景RGB-D-姿态数据用于完整场景覆盖

Jiale Liu, Jungang Li, Jieming Yu, Xinglin Yu, Zihao Dongfang, Zongjian Ding, Kaifeng Ding, Yi Yang, Lidong Chen, Yang Zou, Shunwen Bai, Jiahuan Zhang, Haoran Huang, Shan Huang, Yudong Gao, Mingjun Cheng

机构 * Zhejiang University（浙江大学） ； The Hong Kong University of Science and Technology (Guangzhou)（香港科学与技术大学（广州）） ； The Hong Kong University of Science and Technology（香港科学与技术大学） ； Vorynel（沃伊内尔） ； Xinjiang University（新疆大学） ； Wuhan Polytechnic University（武汉职业技术学院） ； Tianjin University（天津大学） 纠错

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.GR、cs.RO

AI总结 本文提出CM-EVS，通过COVER算法生成稀疏全景RGB-D-姿态数据，实现低冗余且可追溯的完整场景覆盖，提升3D学习的几何一致性。

Comments 35 pages including appendix. Code and dataset: https://github.com/Strange-animalss/CM-EVS

RigidFormer: Learning Rigid Dynamics using Transformers

RigidFormer：基于变换器学习刚体动力学

Zhiyang Dou, Minghao Guo, Haixu Wu, Doug Roble, Tuur Stuyck, Wojciech Matusik

机构 * MIT（麻省理工学院） ； Meta 纠错

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.GR、cs.RO

AI总结 RigidFormer通过对象中心的变换器模型，高效建模无网格表示的高保真刚体动力学，采用可控积分步长和锚点池化技术，实现对点云等无网格输入的高精度模拟。

Comments Project Page: https://people.csail.mit.edu/frankzydou/projects/RigidFormer/index.html

IFG: Internet-Scale Guidance for Functional Grasping Generation

Ray Muxin Liu, Mingxuan Li, Kenneth Shaw, Deepak Pathak

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.GR、cs.RO

Comments Website at https://ifgrasping.github.io/

Physically Based Neural LiDAR Resimulation

Richard Marcus, Marc Stamminger

专题命中 点云 ：novel view synthesis(abstract)；分类 cs.CV、cs.GR、cs.RO

Comments Accepted at ITSC 2025, Gold Coast Australia

ASH: A Modern Framework for Parallel Spatial Hashing in 3D Perception

Wei Dong, Yixing Lao, Michael Kaess, Vladlen Koltun

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.GR、cs.RO

Comments 18 pages, 19 figures

Objaverse: A Universe of Annotated 3D Objects

Matt Deitke, Dustin Schwenk, Jordi Salvador, Luca Weihs, Oscar Michel, Eli VanderBilt, Ludwig Schmidt, Kiana Ehsani, Aniruddha Kembhavi, Ali Farhadi

专题命中 点云 ：3D vision(abstract)；分类 cs.CV、cs.GR、cs.RO

Comments Website: objaverse.allenai.org

Diffusion-based Molecule Generation with Informative Prior Bridges

Lemeng Wu, Chengyue Gong, Xingchao Liu, Mao Ye, Qiang Liu

专题命中 点云 ：point cloud(abstract)；3D generation(abstract)

Fixing Malfunctional Objects With Learned Physical Simulation and Functional Prediction

Yining Hong, Kaichun Mo, Li Yi, Leonidas J. Guibas, Antonio Torralba, Joshua B. Tenenbaum, Chuang Gan

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.GR、cs.RO

Comments CVPR 2022. Project page: http://fixing-malfunctional.csail.mit.edu

DurLAR: A High-fidelity 128-channel LiDAR Dataset with Panoramic Ambient and Reflectivity Imagery for Multi-modal Autonomous Driving Applications

Li Li, Khalid N. Ismail, Hubert P. H. Shum, Toby P. Breckon

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.RO；3D vision(journal_ref)

Comments Accepted by 3DV 2021; 13 pages, 14 figures; Dataset at https://github.com/l1997i/durlar

Journal ref Proc. Int. Conf. on 3D Vision (3DV 2021)

IMU-Free Body-Frame State Estimation with Sparse Scene Flow for Quadcopters

面向四旋翼无人机的基于稀疏场景流的无IMU机体坐标系状态估计

Daniel Grønhaug, Sofie Markeset, Mathias Kolberg

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.RO

AI总结 该研究提出一种无IMU的四旋翼仅视觉状态估计系统，采用扩展卡尔曼滤波与4视图光束平差法，生成机体坐标系状态估计与稀疏场景流，无需GPS等世界坐标系基础设施。

Comments 56 pages, 5 figures, 2 tables. Evaluated on the VID dataset ( arXiv:2103.11152 (https://arxiv.org/abs/2103.11152) )

CVSD-Reg: Cross-Modal Visual Semantic Prior Distillation for Robust LiDAR Registration

CVSD-Reg：用于鲁棒激光雷达配准的跨模态视觉语义先验蒸馏

Eunsoo Im, Junghun Suh, Gyeonggwan Lee, Seunghwan Hong

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.RO

AI总结 本文提出CVSD-Reg框架，通过蒸馏视觉基础模型的语义先验提升激光雷达配准的鲁棒性，在多数据集上的成功率优于现有方法，且无需相机输入或事后ICP精修。

FlexAM: Flexible Appearance-Motion Decomposition for Versatile Video Generation Control

FlexAM: 一种灵活的外观-运动分解方法用于多功能视频生成控制

Mingzhi Sheng, Zekai Gu, Peng Li, Cheng Lin, Hao-Xiang Guo, Ying-Cong Chen, Yuan Liu

机构 * HKUST(GZ)（香港科技大学（广州）） ； HKUST（香港科技大学） ； MUST（澳门大学） ； Tsinghua University（清华大学） 纠错

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.GR

AI总结 FlexAM通过引入新型3D控制信号，实现外观与运动的解耦，提升视频生成任务的灵活性和生成质量。

Comments Codes: https://github.com/IGL-HKUST/FlexAM

LiDAR-based 3D Change Detection at City Scale

基于LiDAR的城市级三维变化检测

Hezam Albaqami, Haitian Wang, Xinyu Wang, Muhammad Ibrahim, Zainy M. Malakan, Abdullah M. Algamdi, Mohammed H. Alghamdi, Ajmal Mian

机构 * Department of Computer Science and Artificial Intelligence, University of Jeddah（计算机科学与人工智能系，朱德亚大学） ； Department of Computer Science and Software Engineering, University of Western Australia（计算机科学与软件工程系，西澳大学） ； Department of Data Science, Umm Al-Qura University（数据科学系，乌姆·阿勒·卡拉大学） ； Department of Information and Technology Systems, College of Computer Science and Engineering, University of Jeddah（信息与技术系统系，计算机科学与工程学院，朱德亚大学） 纠错

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.RO

AI总结 本文提出了一种基于LiDAR的城市级三维变化检测方法，通过多分辨率NDT和ICP对齐数据，结合语义分割和双分匹配优化，实现了高精度的变化检测性能。

Vernata: Self-Supervised Learning of LiDAR Point Representations

Vernata：激光雷达点表示的自监督学习

Oliver Lemke, Alexander Liniger, Abel Gawel, Marco Hutter

机构 * Robotics and AI Institute（机器人与人工智能研究院） ； ETH Zurich（苏黎世联邦理工学院） 纠错

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.RO

AI总结 本研究提出基于Sonata架构的Vernata自监督学习框架，通过三项扩展优化激光雷达点云表示，在多数据集上相比基线取得显著性能提升，模态减少场景下仍保持竞争力

Comments IROS 2026. Implementation: https://github.com/rai-opensource/vernata

Multimodal Plant Root Phenotyping with Integration of 3D Skeleton Extraction and Language Analysis

结合三维骨架提取与语言分析的多模态植物根表型分析

Jiakai Lin, Zijun Li, Guoyu Lu

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.RO

AI总结 该研究提出多模态机器人AI框架，结合W-LBC无监督三维骨架提取与证据优先语言建模，微调GPT实现12种植物根表型的可解释分析，建立了定量与语义结合的根表型统一分析范式。

Intrinsic and Triangulation-Agnostic Attention: A Simple and Powerful Approach for Learning on Meshes

内在与三角剖分无关的注意力：一种用于网格学习的简单而强大的方法

Ashwath Shetty, Zihan Zhu, Soeren Pirk, Noam Aigerman

专题命中 点云 ：point cloud(abstract)；分类 cs.CV、cs.GR

AI总结 该研究为三角形网格学习提出内在与三角剖分无关的注意力机制，通过基于几何处理原理修改注意力机制，用内在网络创建相关量并经有限元离散化设计运行机制，实验表明该方法在多个任务上超当前最优水平。