AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

SpaR3D-MoE: Adaptive 3D Spatial Reasoning from Sparse Views Meets Geometry-Inductive Mixture-of-Experts

Haida Feng, Hao Wei, Haolin Wang, Shiwei Li, Chade Li, Yihong Wu

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 0 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 Adaptive Spatiotemporal Manifold Sampling (vs uniform), 32 frames (ablation)SpaR3D-MoE · VSI-Benchaverage accuracy: 62.4 -> 63.5Improved
add ASMS (vs uniform), 32 frames (ablation)SpaR3D-MoE · VSI-Bench Route Planaccuracy: 39.8 -> 44.0Improved
add ASMS (vs uniform), 8 frames (ablation)SpaR3D-MoE · VSI-Benchaverage accuracy: 56.6 -> 57.8Improved
add ASMS (vs uniform), 16 frames (ablation)SpaR3D-MoE · VSI-Benchaverage accuracy: 61.2 -> 61.9Improved
add holistic expert E0 (ablation)SpaR3D-MoE · VSI-Benchaverage accuracy: 61.8 -> 63.5Improved
add geometric-semantic cross-attention expert E1 (ablation)SpaR3D-MoE · VSI-Benchaverage accuracy: 62.8 -> 63.5Improved
add pose-conditioned dynamic adapter expert E2 (ablation)SpaR3D-MoE · VSI-Benchaverage accuracy: 59.3 -> 63.5Improved
add gravity-aligned structural expert E3 (ablation)SpaR3D-MoE · VSI-Benchaverage accuracy: 62.4 -> 63.5Improved
add camera pose bias to the router (ablation)Base Router · VSI-Benchaverage accuracy: 61.2 -> 62.5Improved
add instruction query guidance to the router (ablation)Base Router · VSI-Benchaverage accuracy: 61.2 -> 61.8Improved
combine pose bias and query guidance in the router (ablation)Base Router · VSI-Benchaverage accuracy: 61.2 -> 63.5Improved
use SpaR3D-MoE instead of Qwen3VL-8BQwen3VL-8B · VSI-Benchaverage accuracy: 55.7 -> 63.5Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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