Robust Rotation Synchronization via Low-rank and Sparse Matrix Decomposition
TLDR
This paper deals with the rotation synchronization problem, which arises in global registration of 3D point-sets and in structure from motion in an unprecedented way as a low-rank and sparse matrix decomposition that handles both outliers and missing data.About:
This article is published in Computer Vision and Image Understanding.The article was published on 2018-09-01 and is currently open access. It has received 58 citations till now. The article focuses on the topics: Sparse matrix & Matrix decomposition.read more
Figures
Figure 2: L is a low rank completion of pX. ![Figure 12: Median angular error [degrees] versus execution time [seconds]. The dashed line connects all the methods which are not dominated by any other method, meaning that no algorithm on this line has both execution time and angular error lower than the others. Only robust algorithms and the largest datasets from (Wilson and Snavely, 2014) are reported.](/figures/figure-12-median-angular-error-degrees-versus-execution-time-1iapaa9t.png)
Figure 12: Median angular error [degrees] versus execution time [seconds]. The dashed line connects all the methods which are not dominated by any other method, meaning that no algorithm on this line has both execution time and angular error lower than the others. Only robust algorithms and the largest datasets from (Wilson and Snavely, 2014) are reported. 
Table 4: Execution times (seconds) of rigid-motion synchronization. The number of point sets and the percentage of missing pairs are also reported. 
Table 3: Mean errors (rotations in degrees, translations in millimetres) on absolute motions for the Stanford and Aim@Shape-Visionair repositories. The number of point sets and the percentage of missing pairs are also reported. The lowest errors are highlighted in boldface. 
Table 5: Cross-sections of registered point-sets. 
Figure 11: ROC curves of outlier detection for di↵erent percentages of outliers, with p “ 0.2. A fixed level of noise is applied to the inlier rotations in this experiment.
![Figure 12: Median angular error [degrees] versus execution time [seconds]. The dashed line connects all the methods which are not dominated by any other method, meaning that no algorithm on this line has both execution time and angular error lower than the others. Only robust algorithms and the largest datasets from (Wilson and Snavely, 2014) are reported.](/figures/figure-12-median-angular-error-degrees-versus-execution-time-1iapaa9t.webp)
Citations
More filters
Journal ArticleDOI
Modeling Perceptual Aliasing in SLAM via Discrete–Continuous Graphical Models
TL;DR: Experimental results on standard benchmarking datasets show that the proposed technique compares favorably with state-of-the-art methods while not relying on an initial guess for optimization.
Proceedings ArticleDOI
Learning Transformation Synchronization
TL;DR: In this paper, a neural network is used to learn the weights associated with each relative transformation, which can then be used to perform transformation synchronization in a two-step process. But the performance of such approaches heavily relies on the quality of the input relative transformations.
Proceedings ArticleDOI
Outlier-Robust Spatial Perception: Hardness, General-Purpose Algorithms, and Guarantees
TL;DR: This paper proposes a simple general-purpose algorithm, named adaptive trimming, to remove outliers, that leverages recently-proposed global solvers that are able to solve outlier-free problems, and iteratively removes measurements with large errors.
Book ChapterDOI
NeuRoRA: Neural Robust Rotation Averaging
TL;DR: This work aims to build a neural network that learns the noise patterns from the data and predict/regress the model parameters from the noisy relative orientations and replaces robust optimization methods by a graph-based network.
Posted Content
Outlier-Robust Estimation: Hardness, Minimally-Tuned Algorithms, and Applications
TL;DR: Two unifying formulations for outlier-robust estimation are introduced, one based on homotopy methods and the other on combinatorial methods, and the first proposes the first minimally tuned algorithms forOutlier rejection, which dynamically decide how to separate inliers from outliers.
References
More filters
Journal ArticleDOI
Distinctive Image Features from Scale-Invariant Keypoints
TL;DR: This paper presents a method for extracting distinctive invariant features from images that can be used to perform reliable matching between different views of an object or scene and can robustly identify objects among clutter and occlusion while achieving near real-time performance.
Journal ArticleDOI
A method for registration of 3-D shapes
Paul J. Besl,H.D. McKay +1 more
TL;DR: In this paper, the authors describe a general-purpose representation-independent method for the accurate and computationally efficient registration of 3D shapes including free-form curves and surfaces, based on the iterative closest point (ICP) algorithm, which requires only a procedure to find the closest point on a geometric entity to a given point.
Book
Distributed Optimization and Statistical Learning Via the Alternating Direction Method of Multipliers
TL;DR: It is argued that the alternating direction method of multipliers is well suited to distributed convex optimization, and in particular to large-scale problems arising in statistics, machine learning, and related areas.
Journal ArticleDOI
A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
Amir Beck,Marc Teboulle +1 more
TL;DR: A new fast iterative shrinkage-thresholding algorithm (FISTA) which preserves the computational simplicity of ISTA but with a global rate of convergence which is proven to be significantly better, both theoretically and practically.
Journal ArticleDOI
De-noising by soft-thresholding
TL;DR: The authors prove two results about this type of estimator that are unprecedented in several ways: with high probability f/spl circ/*/sub n/ is at least as smooth as f, in any of a wide variety of smoothness measures.