AI Paper Atlas

Research map / Point clouds / Registration and matching

Certifiably Optimal Outlier-Robust Geometric Perception: Semidefinite Relaxations and Scalable Global Optimization

Heng Yang, Luca Carlone

TPAMI 2022 · Point clouds › Registration and matching · approach: Quantum and certifiable correspondence · 58 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use sparse monomial basis instead of dense Lasserre basis (ablation)TLS robust estimation SDP relaxation · Rotation averaging SDP size (d=9, N=200)SDP size (n1 and moment constraints): n1 > 20,000 -> n1 approximately 2,000; moment constraints > 100,000,000 -> approximately 1,000,000Improved
use sparse relaxation instead of dense hierarchy (ablation)TLS robust estimation SDP relaxation · Estimation with fewer than 20 measurementssolver tractability: 100 times reduction in the size of the SDPImproved
solve sparse SDP with off-the-shelf solver (MOSEK)Sparse semidefinite relaxation · Small problems, 30 measurementsscalability: 30 measurementsNo clear effect
replace off-the-shelf solvers with the STRIDE solverSparse SDP relaxation · Medium-scale problemssolve time: up to 100 times fasterImproved
bootstrap STRIDE with RANSAC or GNC heuristicsSTRIDE · Six geometric perception problemsglobal optimality certificationImproved
apply sparse relaxation with up to 60%-90% outliersSparse semidefinite relaxation · Six geometric perception problemsexactness of relaxation: 60%-90% outliersImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy