AI Paper Atlas

Research map / 3D vision / Human pose estimation

MVHumanNet++: A Large-Scale Dataset of Multi-View Daily Dressing Human Captures With Richer Annotations for 3D Human Digitization

Chenghong Li, Hongjie Liao, Yihao Zhi, Xihe Yang, Zhengwentai Sun, Jiahao Chang, Shuguang Robert Cui, Xiaoguang Han

TPAMI 2026 · 3D vision › Human pose estimation · approach: Human shape and pose regression · 11 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
train on 8 views instead of 1 viewFR-Head · MVHumanNet++ action recognitiontop-1 accuracy (%): 30.25 -> 78.19Improved
train on 8 views instead of 1 viewInfoGCN · MVHumanNet++ action recognitiontop-1 accuracy (%): 25.23 -> 76.55Improved
scale NeRF training outfits from 100 to 5000IBRNet · MVHumanNet++ generalizable NeRFPSNR: 26.05 -> 29.00Improved
scale NeRF training outfits from 100 to 5000GPNeRF · MVHumanNet++ generalizable NeRFPSNR: 23.27 -> 24.69Improved
pretrain on MVHumanNet++ then finetune on HuMManIBRNet · HuMMan testPSNR: 28.06 -> 29.46Improved
pretrain on MVHumanNet++ then finetune on HuMManGPNeRF · HuMMan testPSNR: 20.95 -> 21.89Improved
use MVHumanNet++ pretrained model without finetuningIBRNet · HuMMan testPSNR: 28.06 -> 27.48Hurt
use improved SMPLX annotations from MVHumanNet++ (ablation)Animatable Gaussians · MVHumanNet++ per-subject 3DGSPSNR: 27.259 -> 28.593Improved
scale LaRa training outfits from 100 to 5000LaRa · MVHumanNet++ generalizable 3DGSPSNR: 21.431 -> 22.441Improved
scale EVA-Gaussian training outfits from 100 to 5000EVA-Gaussian · MVHumanNet++ generalizable 3DGSPSNR: 26.984 -> 28.544Improved
scale training subjects from 3000 to 5500StyleGAN2 · MVHumanNet++ generationFID: 14.05 -> 7.08Improved
scale training subjects from 3000 to 5500GET3D · MVHumanNet++ generationFID: 41.54 -> 25.12Improved

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