Article
Computer Science, Artificial Intelligence
Shubham Tulsiani, Tinghui Zhou, Alexei A. Efros, Jitendra Malik
Summary: We study the consistency between a 3D shape and a 2D observation, and propose a differentiable formulation to compute the gradients of the 3D shape given an observation from any view. Our method uses a differentiable ray consistency term and can incorporate various types of multi-view observations as supervision for learning single-view 3D prediction. We provide empirical analysis and show improvement over existing techniques for single-view object reconstruction.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Computer Science, Artificial Intelligence
Xuancheng Zhang, Rui Ma, Changqing Zou, Minghao Zhang, Xibin Zhao, Yue Gao
Summary: Reconstructing 3D shape from a single-view image using deep learning has gained popularity, but existing methods suffer from the lack of explicit structure modeling and loss of view information. In this paper, we propose VGSNet, an encoder-decoder architecture that jointly learns the feature representation of 2D image and 3D shape to achieve geometry and structure reconstruction from a single-view image.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Computer Science, Artificial Intelligence
Qichuan Geng, Hong Zhang, Feixiang Lu, Xinyu Huang, Sen Wang, Zhong Zhou, Ruigang Yang
Summary: This paper introduces the first part-aware approach for joint part-level car parsing and reconstruction in single street view images. By incorporating dense part information and a class-consistent method, significant improvements in part segmentation performance on real street views are achieved, leading to state-of-the-art pose and shape estimation results on the ApolloCar3D dataset.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Engineering, Electrical & Electronic
Yuxuan Chen, Ben Wang, Qiongwei Li, Yujun Zhong, Yi Jin, Changan Zhu
Summary: This article proposes a FOV-enlarged single-camera 3-D shape reconstruction system by using a saccade mirror to generate virtual cameras and reconstruct objects with multiview images. The system achieves a larger FOV without sacrificing image resolution, simplifies the calibration process, and demonstrates robustness in real-life experiments.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Computer Science, Information Systems
Bilal Ahmad, Pal Anders Floor, Ivar Farup, Oistein Hovde
Summary: Capsule endoscopy is emerging as an alternative to traditional colonoscopy, utilizing a wireless camera to examine the gastrointestinal tract. The study focuses on reconstructing 3D shapes from wireless capsule endoscopy (WCE) images to provide enhanced viewing for gastroenterologists. Different methods, such as shape from shading (SFS) and shape from focus (SFF), are evaluated and compared. The results show that both methods successfully reconstruct the 3D shapes of colon images, but SFF performs better in retaining details. Additionally, the subjective experiments indicate that contrast enhanced images and 3D models are preferred and considered useful by gastroenterologists.
Article
Automation & Control Systems
Rongshan Chen, Yuancheng Yang, Chao Tong
Summary: In this paper, the authors propose a graph-based implicit function G2IFu, which successfully reconstructs a highly detailed 3D object mesh from a single image. By mapping graphs to implicit values, G2IFu improves the prediction accuracy of implicit functions compared to traditional point-based methods. Additionally, the authors introduce prior boundary loss and self-attention module to enhance the performance of G2IFu.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Software Engineering
Kestutis Karciauskas, Jorg Peters
Summary: This article discusses the problem of filling multi-sided holes and introduces a new algorithm to improve surface shape. The new algorithm addresses the issues of the old algorithms and emphasizes the balance between smoothness and good shape.
COMPUTER-AIDED DESIGN
(2023)
Article
Mechanics
Chaoyue Gong, Yuchen Song, Guangyuan Huang, Wuguang Chen, Junlian Yin, Dezhong Wang
Summary: This paper proposes a novel method for the 3D reconstruction of bubble shape based on single-view images. By utilizing grayscale information and neural networks, it can automatically reconstruct the rough 3D shapes of one side of bubbles in the images.
INTERNATIONAL JOURNAL OF MULTIPHASE FLOW
(2022)
Article
Computer Science, Software Engineering
Aihua Mao, Canglan Dai, Qing Liu, Jie Yang, Lin Gao, Ying He, Yong-Jin Liu
Summary: In this article, a novel approach named STD-Net is proposed for 3D reconstruction using mesh representation suitable for characterizing complex structures and geometry details. The method includes an auto-encoder network for recovering the object structure from a single-view image, a topology-adaptive GCN for updating vertex position for meshes of complex topology, and a unified mesh deformation block for deforming the structural boxes into structure-aware meshes. Evaluation on ShapeNet and PartNet demonstrates that STD-Net outperforms state-of-the-art methods in reconstructing complex structures and fine geometric details.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2023)
Article
Computer Science, Software Engineering
Xiuping Liu, Hua Huang, Weiming Wang, Jun Zhou
Summary: This work introduces a neural network model based on multi-view representations for learning style transformation while preserving contents of 3D shapes. By learning style transformation between different style domains, the model preserves structural details of 3D shapes and outperforms baselines and state-of-the-art approaches in experiments.
Article
Computer Science, Artificial Intelligence
Yang Zhou, Hanjie Wu, Wenxi Liu, Zheng Xiong, Jing Qin, Shengfeng He
Summary: Synthesizing novel views from a single view image is a challenging task, but can be improved by expanding to a multi-view setting. By leveraging stereo prior, a pseudo-stereo viewpoint is generated to assist in 3D reconstruction, making the view synthesis process simpler. A self-rectified stereo synthesis approach is proposed to correct erroneous regions and generate high-quality stereo images.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2023)
Article
Engineering, Multidisciplinary
Chuang Lu, Leilei Chen, Jinling Luo, Haibo Chen
Summary: This study proposes a new acoustic shape optimization method by combining isogeometric subdivision surfaces with the boundary element method. The geometry of the structural boundary is constructed using Catmull-Clark subdivision surfaces and the sound propagation in the unbounded domain is simulated using the boundary element method. Compared to conventional methods, this approach eliminates jagged geometry by using a multiresolution approach without the need for complex meshing and volume parameterization.
ENGINEERING ANALYSIS WITH BOUNDARY ELEMENTS
(2023)
Article
Geosciences, Multidisciplinary
Mohammad Moulaeifard, Simon Bernard, Florian Wellmann
Summary: This work proposes a flexible framework for creating and interacting with geological models using explicit surface representations. The control mesh and semi-sharp-crease values are determined to enable the representation of complex structural settings with a low number of control points. The method is combined with a particle swarm optimization approach for automatic optimization of vertex position and crease sharpness values.
GEOSCIENTIFIC MODEL DEVELOPMENT
(2023)
Article
Computer Science, Artificial Intelligence
Jinzhi Zhang, Mengqi Ji, Guangyu Wang, Zhiwei Xue, Shengjin Wang, Lu Fang
Summary: This paper proposes SurRF, an unsupervised multi-view stereopsis pipeline that learns Surface Radiance Field. By defining the radiance field on a continuous and explicit 2D surface, SurRF provides a compact representation while maintaining complete shape and realistic texture, leading to competitive results for large-scale complex scenes.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Environmental Sciences
Guoqing Gao, Liang Yang, Quan Zhang, Chongmin Wang, Hua Bao, Changhui Rao
Summary: The paper proposes a two-stage approach for 3D shape reconstruction by fusing different features to improve accuracy, outperforming state-of-the-art methods.
Article
Computer Science, Software Engineering
Jiri Kosinka, Thomas J. Cashman
COMPUTER AIDED GEOMETRIC DESIGN
(2015)
Article
Computer Science, Software Engineering
Jonathan Taylor, Lucas Bordeaux, Thomas Cashman, Bob Corish, Cem Keskin, Toby Sharp, Eduardo Soto, David Sweeney, Julien Valentin, Benjamin Luff, Arran Topalian, Erroll Wood, Sameh Khamis, Pushmeet Kohli, Shahram Izadi, Richard Banks, Andrew Fitzgibbon, Jamie Shotton
ACM TRANSACTIONS ON GRAPHICS
(2016)
Article
Computer Science, Software Engineering
Thomas J. Cashman, Ursula H. Augsdoerfer, Neil A. Dodgson, Malcolm A. Sabin
ACM TRANSACTIONS ON GRAPHICS
(2009)
Article
Computer Science, Software Engineering
Thomas J. Cashman, Kai Hormann
COMPUTER GRAPHICS FORUM
(2012)
Proceedings Paper
Computer Science, Artificial Intelligence
Mariano Jaimez, Thomas J. Cashman, Andrew Fitzgibbon, Javier Gonzalez-Jimenez, Daniel Cremers
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017)
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
David Joseph Tan, Thomas Cashman, Jonathan Taylor, Andrew Fitzgibbon, Daniel Tarlow, Sameh Khamis, Shahram Izadi, Jamie Shotton
2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)
(2016)