AI Paper Atlas

Research map / 3D vision / Human pose estimation

Adaptive Multi-View and Temporal Fusing Transformer for 3D Human Pose Estimation

Hui Shuai, Lele Wu, Qingshan Liu

TPAMI 2022 · 3D vision › Human pose estimation · approach: Transformer-based multi-view 3D pose · 99 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 Confidence Attentive Aggregation instead of no confidence (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 29.4 -> 28.0Improved
concatenate 2D pose with confidence instead of ignoring it (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 29.4 -> 29.2Improved
use CAA instead of concatenating confidence (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 29.2 -> 28.0Improved
use MFT instead of conventional transformer for multi-view fusion (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 31.9 -> 28.0Improved
use MFT instead of point transformer (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 28.4 -> 28.0Improved
add transform matrix Tij to MFT (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 30.5 -> 28.0Improved
add MFT module (train on 2 views, test on 4 views) (ablation)MTF-Transformer · Human3.6M generalizationMPJPE: 57.0 -> 39.0Improved
use random block mask rate 0.4 instead of 0 (ablation)MTF-Transformer · Human3.6M, mean over 1-4 viewsMPJPE: 89.0 -> 34.9Improved
use mask rate 1.0 instead of 0.4 (ablation)MTF-Transformer · Human3.6M, mean over 1-4 viewsMPJPE: 34.9 -> 107.2Hurt
add one synthesized view in training instead of none (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 28.4 -> 28.0Improved
add a second synthesized view (2 instead of 1) (ablation)MTF-Transformer · Human3.6M, 4 views, T=1MPJPE: 28.0 -> 28.1Hurt
increase test sequence length from 1 to 27 frames (ablation)MTF-Transformer · Human3.6M, 4 viewsMPJPE: 27.46 -> 26.21Improved

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