AI Paper Atlas

Research map / 3D vision / Depth and stereo

SMVRT: Implicit Human 3D Modeling Using Sparse Multi-View Volumetric Reconstruction with Transformer Fusion

Chuanmao Fan, Chenxi Zhao, Ye Duan

CVPR 2026 · 3D vision › Depth and stereo · approach: Cost-volume multi-view stereo

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 MMFT multi-view multi-level 2D-to-3D fusion transformer (ablation)SMVRT no-fusion baseline · THUman2.1Chamfer L1 mean: 4.40 -> 3.84Improved
add QFT query point fusion transformer (ablation)SMVRT no-fusion baseline · THUman2.1Chamfer L1 mean: 4.40 -> 4.02Improved
add AFM alternating global-local 2D fusion (ablation)SMVRT no-fusion baseline · THUman2.1Chamfer L1 mean: 4.40 -> 4.37No clear effect
add AFM on top of MMFT+QFT (ablation)SMVRT · THUman2.1Chamfer L1 mean: 3.61 -> 3.58No clear effect
combine MMFT and QFT (ablation)SMVRT · THUman2.1Chamfer L1 mean: 3.84 -> 3.61Improved
increase input views from 4 to 8 (ablation)SMVRT · THUman2.1Chamfer L1 mean: 3.58 -> 2.96Improved
reduce input views from 4 to 2 (ablation)SMVRT · THUman2.1Chamfer L1 mean: 3.58 -> 8.74Hurt
raise image resolution to 1024 with 8 views (ablation)SMVRT · THUman2.1Chamfer L1 mean: 2.96 -> 2.91No clear effect
add Sapiens normal maps as input (ablation)SMVRT · THUman2.1Chamfer L1 mean: 3.90 -> 3.71Improved
add depth, then Plucker rays, to the inputs (ablation)SMVRT · THUman2.1Chamfer L1 mean: 3.71 -> 3.65 -> 3.58Improved
use Mamba instead of Transformer blocks in AFM (ablation)SMVRT · THUman2.1Chamfer L1 mean: 3.577 -> 3.586No clear effect
add low-level 2D detail features at query points (ablation)QFT · THUman2.1Chamfer L1 mean: 3.92 -> 3.58Improved

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