AI Paper Atlas

Research map / 3D vision / 3D generation

MVGGT: Multimodal Visual Geometry Grounded Transformer for Multiview 3D Referring Expression Segmentation

Changli Wu, Haodong Wang, Jiayi Ji, Yutian Yao, Chunsai Du, Jihua Kang, Yanwei Fu, Liujuan Cao

CVPR 2026 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images · 6 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 PVSO (ablation)MVGGT without multimodal branch · MVRefer overallmIoU global: 26.9 -> 32.0Improved
add PVSO (ablation)MVGGT without multimodal branch · MVRefer overallmIoU view: 41.1 -> 47.5Improved
add the MVGGT multimodal branch with language guidance (ablation)MVGGT with PVSO · MVRefer overallmIoU global: 32.0 -> 39.9Improved
add the MVGGT multimodal branch with language guidance (ablation)MVGGT with PVSO · MVRefer hardmIoU global: 19.0 -> 24.4Improved
add no-target suppression to random sampling (ablation)PVSO · MVRefer overallmIoU global: 32.4 -> 36.7Improved
use hybrid sampling with no-target ratio 0.5 instead of 0.25 (ablation)PVSO · MVRefer overallmIoU global: 35.7 -> 39.9Improved
raise no-target ratio from 0.5 to 0.75 (ablation)PVSO · MVRefer overallmIoU global: 39.9 -> 25.7Hurt
use late fusion instead of early fusion of language (ablation)MVGGT · MVRefer overallmIoU global: 36.1 -> 39.9Improved
use late fusion instead of middle fusion of language (ablation)MVGGT · MVRefer overallmIoU global: 37.5 -> 39.9Improved
use 12 multimodal layers instead of 6 (ablation)MVGGT · MVRefer overallmIoU global: 37.3 -> 39.9Improved
expand multimodal branch from 12 to 16 layers (ablation)MVGGT · MVRefer overallmIoU global: 39.9 -> 25.4Hurt
use end-to-end MVGGT instead of two-stage reconstruct-then-segmentMVGGT vs two-stage · MVRefer overallmIoU global: 18.5 -> 39.9Improved

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