AI Paper Atlas

Research map / Point clouds / Point cloud classification

Fast and Interpretable 2D Homography Decomposition: Similarity-Kernel-Similarity and Affine-Core-Affine Transformations

Shen Cai, Zhanhao Wu, Lingxi Guo, Jiachun Wang, Siyu Zhang, Junchi Yan, Shuhan Shen

TPAMI 2025 · Point clouds › Point cloud classification · approach: Equivariant point cloud networks · 6 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 SKS decomposition instead of NDLT-SVD4-point homography solver · 4-point homographyFLOPs: about 162x speedup (27400 -> 169)Improved
use SKS instead of HO-SVD4-point homography solver · 4-point homographyFLOPs: about 11x (1800 -> 169)Improved
use SKS instead of GPT-LU4-point homography solver · 4-point homographyFLOPs: about 12x (1950 -> 169)Improved
use SKS instead of RHO-GE4-point homography solver · 4-point homographyFLOPs: about 1.3x (223 -> 169)Improved
use ACA instead of SKS4-point homography solver · 4-point homographyFLOPs: 169 -> 97Improved
use ACA instead of NDLT-SVD, HO-SVD, GPT-LU, RHO-GE4-point homography solver · 4-point homographyFLOPs fraction: 0.35%, 5.4%, 5.0% and 43% of their FLOPsImproved
use ACA(*) instead of RHO-GE(*) at O14-point homography · CPU runtime O1runtime (us): 0.0327 -> 0.0146Improved
use ACA(*) instead of NDLT-SVD(**) at O24-point homography · CPU runtime O2runtime (us): 12.5 -> 0.0145Improved
use SKS(*) instead of RHO-GE(*) at O24-point homography · CPU runtime O2runtime (us): 0.0287 -> 0.0252No clear effect
use single instead of double precision for SKS at Od (ablation)SKS · CPU runtime Odruntime (us): 0.0901 -> 0.0595Improved
specialize ACA to a rectangular source plane (TensorACA)ACA · rectangle sourceFLOPs: 97 -> 47Improved
drop pivoting as in RHO-GERHO-GE · 4-point homography robustnessnumerical robustness: fails for improper point orderHurt

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