AI Paper Atlas

Research map / 3D vision / Human pose estimation

Deep Autoencoder for Combined Human Pose Estimation and Body Model Upscaling

Matthew Trumble, Andrew Gilbert, Adrian Hilton, John Collomosse

ECCV 2018 · 3D vision › Human pose estimation · approach: Monocular 2D-to-3D pose lifting · 62 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 the decoder and volume reconstruction loss (dual loss) to the pose encoder (ablation)pose autoencoder · TotalCapturemean joint error (mm): 85.4 -> 38.6Improved
increase upscaling from x1 to x2AutoEnc-LSTM · TotalCapturemean joint error (mm): 38.6 -> 34.1Improved
increase upscaling from x2 to x4AutoEnc-LSTM · TotalCapturemean joint error (mm): 34.1 -> 35.5No clear effect
increase upscaling from x2 to x4AutoEnc · TotalCapturevoxel MSE x10^-3: 0.37 -> 0.56No clear effect
replace tricubic upsampling with the autoencoder at x2AutoEnc-x2 · TotalCapturevoxel MSE x10^-3: 9.88 -> 0.37Improved
replace tricubic upsampling with the autoencoder at x4AutoEnc-x4 · TotalCapturevoxel MSE x10^-3: 10.56 -> 0.56Improved
use the proposed autoencoder instead of the previous best (Martinez)proposed method · Human3.6M (S9,S11) meanmean joint error (mm): 62.9 -> 62.5Improved
use the proposed autoencoder instead of Trumble's earlier CNNAutoEnc-x2-LSTM · TotalCapture meanmean joint error (mm): 70.0 -> 34.1Improved
evaluate on other subject pairs instead of S9/S11proposed method · Human3.6M (S9,S11)mean joint error (mm): 62.5 -> 70.2No clear effect
raise or lower the pose loss weight lambda outside the working range (ablation)dual loss training · Human3.6M (S9,S11)convergence / latent pose qualityHurt

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