Multicore bundle adjustment

By Changchang Wu, Sameer Agarwal, Brian Curless, S. M. Seitz. 2011

869 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

RayNet: Learning Volumetric 3D Reconstruction with Ray Potentials (2018, methodology): “The input to our approach is a set of images and their corresponding camera poses, which are obtained using structure-from-motion [36].”
Generic 3D Representation via Pose Estimation and Matching (2016, methodology): “Pose Estimation Baselines: We compared against Structure-from-Motion (visualSfM [59,60] with default components and tuned hyper-parameters for pair-wise pose estimation on 192 × 192 patches and full images) and LIBVISO2 Visual Simo - Serra ( pre-trained ) m= 2.0 ×10 -3 F r e qu e n c y ( % )…”
3D Object Representations for Fine-Grained Categorization (2013, methodology): “We run BB-3D-L for all pairs of images of a category to obtain putative correspondences, and feed feature locations and correspondences into VisualSFM, a front-end for multi-core bundle adjustment [31].”
Image-Based Localization Using LSTMs for Structured Feature Correlation (2016, methodology): “Since none of the datasets provides both an SfM model and the SIFT descriptors, we constructed such models from scratch using VisualSFM [56, 57] and Colmap [44] and registered them against the ground truth poses of the training images.”
Semantic Multi-view Stereo: Jointly Estimating Objects and Voxels (2017, methodology): “As input we assume a set of images and camera poses, which we obtain using structure-from-motion [35, 36].”
Image Retrieval for Image-Based Localization Revisited (2012, methodology): “An SfM reconstruction of the scene was obtained using the freely available Bundler software [31, 39] and SiftGPU [38].”
Reconstructing building mass models from UAV images (2016, neutral): “Most recently, state-19 of-the-art Structure from Motion (SfM) and Multi-View Stereo (MVS) methods [5, 6, 7] have produced extremely compelling results on a wide variety of scenes.”
Tanks and temples (2017, methodology): “VisualSFM [Wu 2011] is a highly optimized incremental SfM pipeline that integrates ideas from multiple projects [Wu 2013; Wu et al. 2011].”
Robust monocular SLAM in dynamic environments (2013, background): “In recent years, lots of optimization methods [11, 40] have been proposed to accelerate BA, where local bundle adjustment involving tens of keyframes and hundreds of 3D points has been proved to be solved at interactive rate [22].”
The use of close-range photogrammetry in zooarchaeology: Creating accurate 3D models of wolf crania to study dog domestication (2016, methodology): “…Lowe, 1999; Wu, 2007), tie-points matching (matching points between pairs of photographs), sparse 3D points generation (Sholts et al., 2011;Wu, 2013;Wu et al., 2011) andfinally to reconstruct high-resolution 3D point-clouds by dense correlation through CMVSPMVS algorithms (i.e. Clustering Views…”
Towards Linear-Time Incremental Structure from Motion (2013, methodology): “Recently, the multicore bundle adjustment [16] takes one step further by using implicit multiplication of the Hessian matrices and Schur Complements, which requires to construct only the O(n) space Jacobian matrices.”
Global, Dense Multiscale Reconstruction for a Billion Points (2015, methodology): “In particular, we used the VisualSfM software [21, 20] to compute camera param-”
Global Motion Estimation from Point Matches (2012, methodology): “0.0072 0.112 0.0044 VisualSFM [24] 0.0099 0.116 0.0046 Sinha et al. [9] 0.1317 – – Martinec [10] (on R25) 0.0153 – – Location Viewpoint Rotation Herz-Jesu-P25 (meters) (degrees) (Frobenius) Our method (GT cal.”
Campus3D: A Photogrammetry Point Cloud Benchmark for Hierarchical Understanding of Outdoor Scene (2020, background): “of Structure from Motion (SfM), Multi-View Stereo (MVS) and UAV techniques [11, 39], photogrammetry point cloud is easily accessible from unmanned aerial vehicle (UAV) imagery.”
A Survey on Structure from Motion (2017, methodology): “A highly eﬃcient multicore CPU and GPU implementation of the preconditioned conjugate gradients (CG) solver (similar to [1], cf. § 4) for bundle adjustment, named multicore bundle adjustment , was introduced in [86].”