AI Paper Atlas

Research map / 3D vision / Human pose estimation

H3WB: Human3.6M 3D WholeBody Dataset and Benchmark

Yue Zhu, Nermin Samet, David Picard

ICCV 2023 · 3D vision › Human pose estimation · approach: Human shape and pose regression · 30 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 diffusion-based 2D refinement after the completion network (ablation)H3WB annotation pipeline · manual cross-check on 600 images3D error all (mm): 29.31 -> 16.98Improved
add completion network to fill missing keypoints (ablation)H3WB annotation pipeline · manual cross-check on 600 images3D error all (mm): 14.87 -> 29.31Hurt
train on H3WB + T3WB instead of H3WB onlySimpleBaseline · H3WB 2D->3DMPJPE all (mm): 125.4 -> 115.3Improved
train on H3WB + T3WB instead of H3WB onlyJointformer · H3WB 2D->3DMPJPE all (mm): 88.3 -> 81.5Improved
train on H3WB + T3WB instead of H3WB onlyJointformer · H3WB I2D->3DMPJPE all (mm): 109.2 -> 84.2Improved
train on H3WB + T3WB instead of H3WB onlySimpleBaseline · H3WB I2D->3DMPJPE all (mm): 268.8 -> 260.5No clear effect
enlarge SimpleBaseline from 6 to 8 layers (Large SimpleBaseline) (ablation)SimpleBaseline · H3WB 2D->3DMPJPE all (mm): 125.4 -> 112.3Improved
remove batch normalization / enlarge SimpleBaseline for incomplete input (ablation)SimpleBaseline · H3WB I2D->3DMPJPE all (mm): 268.8 -> 131.4Improved
add 3D supervision with fixed camera views to CanonPose (ablation)CanonPose · H3WB 2D->3DMPJPE all (mm): 186.7 -> 117.7Improved
use discriminative lifting rather than parametric model fitting (SMPLify-X)2D->3D lifting · H3WB 2D->3DMPJPE all (mm): 188.9 -> 88.3 (Jointformer)Improved
use a single-stage ResNet50 instead of two-stage SHN + SimpleBaselineRGB->3D · H3WB RGB->3DMPJPE all (mm): 182.5 -> 166.7Improved
use part-disjoint DOPE model rather than jointly trained whole-body modelsRGB->3D · H3WB RGB->3DMPJPE all (mm): 132.6 -> 191.3Hurt

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