On the convergence of IRLS and its variants in outlier-robust estimation

L Peng, C Kümmerle, R Vidal - Proceedings of the IEEE …, 2023 - openaccess.thecvf.com
Outlier-robust estimation involves estimating some parameters (eg, 3D rotations) from data
samples in the presence of outliers, and is typically formulated as a non-convex and non …

Scalable 3d registration via truncated entry-wise absolute residuals

T Huang, L Peng, R Vidal… - Proceedings of the IEEE …, 2024 - openaccess.thecvf.com
Given an input set of 3D point pairs the goal of outlier-robust 3D registration is to compute
some rotation and translation that align as many point pairs as possible. This is an important …

Essential matrix estimation using convex relaxations in orthogonal space

A Karimian, R Tron - Proceedings of the IEEE/CVF …, 2023 - openaccess.thecvf.com
We introduce a novel method to estimate the essential matrix for two-view Structure from
Motion (SfM). We show that every 3 by 3 essential matrix can be embedded in a 4 by 4 …

Imbalanced mixed linear regression

P Zilber, B Nadler - Advances in Neural Information …, 2023 - proceedings.neurips.cc
We consider the problem of mixed linear regression (MLR), where each observed sample
belongs to one of $ K $ unknown linear models. In practical applications, the mixture of the …

Block coordinate descent on smooth manifolds: Convergence theory and twenty-one examples

L Peng, R Vidal - arxiv preprint arxiv:2305.14744, 2023 - arxiv.org
Block coordinate descent is an optimization paradigm that iteratively updates one block of
variables at a time, making it quite amenable to big data applications due to its scalability …

Unlabeled principal component analysis

Y Yao, L Peng, M Tsakiris - Advances in Neural Information …, 2021 - proceedings.neurips.cc
We introduce robust principal component analysis from a data matrix in which the entries of
its columns have been corrupted by permutations, termed Unlabeled Principal Component …

Unlabeled Principal Component Analysis and Matrix Completion

Y Yao, L Peng, MC Tsakiris - Journal of Machine Learning Research, 2024 - jmlr.org
We introduce robust principal component analysis from a data matrix in which the entries of
its columns have been corrupted by permutations, termed Unlabeled Principal Component …

Recovering simultaneously structured data via non-convex iteratively reweighted least squares

C Kümmerle, J Maly - Advances in Neural Information …, 2024 - proceedings.neurips.cc
We propose a new algorithm for the problem of recovering data that adheres to multiple,
heterogenous low-dimensional structures from linear observations. Focussing on data …

Sample-Efficient Geometry Reconstruction from Euclidean Distances using Non-Convex Optimization

I Ghosh, A Tasissa, C Kümmerle - arxiv preprint arxiv:2410.16982, 2024 - arxiv.org
The problem of finding suitable point embedding or geometric configurations given only
Euclidean distance information of point pairs arises both as a core task and as a sub …

Efficient and Robust Point Cloud Registration via Heuristics-guided Parameter Search

T Huang, H Li, L Peng, Y Liu… - IEEE Transactions on …, 2024 - ieeexplore.ieee.org
Estimating the rigid transformation with 6 degrees of freedom based on a putative 3D
correspondence set is a crucial procedure in point cloud registration. Existing …