AI Paper Atlas

Research map / 3D vision / Human pose estimation

Scalable and Generalizable Correspondence Pruning via Geometry-Consistent Pre-Training

Tangfei Liao, Xiaoqin Zhang, Tao Wang, Hao Ye, Min Li, Guobao Xiao, Mang Ye

TPAMI 2026 · 3D vision › Human pose estimation · approach: Learned camera pose estimation · 1 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
add local stream to global-only encoder (ablation)CorrFormer · YFCC100M outdoorPose AUC@5: 34.64 -> 37.69Improved
add global stream to local-only encoder (ablation)CorrFormer · YFCC100M outdoorPose AUC@5: 28.65 -> 37.69Improved
add consensus interaction between streams (vs local-to-global without interaction) (ablation)CorrFormer · YFCC100M outdoorPose AUC@5: 36.00 -> 37.69Improved
use implicit weight-shared interaction instead of cross-attention (ablation)CorrFormer · YFCC100M outdoorPose AUC@5: 36.92 -> 37.69Improved
use 60% random masking instead of 40% (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 38.84 -> 39.54Improved
use 60% random masking instead of 80% (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 38.25 -> 39.54Improved
use random masking instead of block masking (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 38.75 -> 39.54Improved
add keypoint distribution loss (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 39.07 -> 39.54Improved
use ordered positional prompts instead of shuffled (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 38.02 -> 39.54Improved
add mask embeddings to positional prompts (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 39.05 -> 39.54Improved
pre-train on SIFT+NN correspondences with 89.23% outliers instead of ground-truth inliers (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 39.54 -> 37.73Hurt
pre-train on DKM+RANSAC correspondences (6.10% outliers) instead of GT inliers (ablation)CorrMAE · YFCC100M outdoorPose AUC@5: 39.54 -> 39.75No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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