AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

3DSRBench: A Comprehensive 3D Spatial Reasoning Benchmark

Wufei Ma, Haoyu Chen, Guofeng Zhang, Yu-Cheng Chou, Jieneng Chen, Celso de Melo, Alan Yuille

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 92 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 DINOv2 as a second vision encoder (ablation)LLaVA-v1.5-7B · 3DSRBenchmean accuracy: 36.8 -> 37.2Improved
add MAE as a second vision encoder (ablation)LLaVA-v1.5-7B · 3DSRBenchmean accuracy: 36.8 -> 33.1Hurt
add SAM as a second vision encoder (ablation)LLaVA-v1.5-7B · 3DSRBenchmean accuracy: 36.8 -> 27.9Hurt
add SAM as a second vision encoder (ablation)LLaVA-v1.5-7B · 3DSRBench heightheight accuracy: 38.5 -> 44.6Improved
replace MLP with SVA connector for CLIP+DINOv2 (ablation)LLaVA-v1.5-7B · 3DSRBenchmean accuracy: 37.2 -> 37.8Improved
replace MLP with SVA connector for CLIP+MAE (ablation)LLaVA-v1.5-7B · 3DSRBenchmean accuracy: 33.1 -> 34.1Improved
scale language model size from 0.5B to 72B (ablation)Qwen2.5 and InternVL2.5 · 3DSRBench-realaccuracyImproved
use proprietary LMMs instead of open-sourced onesproprietary LMMs · 3DSRBench-realoverall accuracy: 44.2 vs 50.9No clear effect
add depth input to SpatialRGPT (ablation)SpatialRGPT · 3DSRBench-realoverall accuracy: 32.7 -> 48.4Improved
add explicit thinking to Gemini-2.0-Flash (ablation)Gemini-2.0-Flash · 3DSRBench-realoverall accuracy: 49.8 -> 51.1Improved
add 3D bounding-box prompting to Gemini-2.0-Flash (ablation)Gemini-2.0-Flash · 3DSRBench-realoverall accuracy: 49.8 -> 47.5Hurt
change camera viewpoint from common to uncommonGPT-4o · 3DSRBench-syntheticoverall accuracy: 51.2 -> 44.3Hurt

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