AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

PanoGrounder: Bridging 2D and 3D with Panoramic Scene Representations for VLM-based 3D Visual Grounding

Seongmin Jung, Seongho Choi, Gunwoo Jeon, Minsu Cho, Jongwoo Lim

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 2 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 EMD loss on digit tokens (ablation)PanoGrounder · ScanReferAcc@0.25: 57.5 -> 58.4; 59.3 -> 60.2Improved
add multi-view fused semantic features (ablation)PanoGrounder · ScanReferAcc@0.25: 58.4 -> 60.4Improved
add geometric range features and geometric QA (ablation)PanoGrounder · ScanReferAcc@0.25: 60.4 -> 61.0Improved
add geometric QA auxiliary training (ablation)PanoGrounder · ScanReferAcc@0.25: 60.2 -> 61.0Improved
use structure-aware camera placement (ray coverage, surface distance, trajectory) vs random (ablation)PanoGrounder · ScanReferAcc@0.25: 51.9 -> 61.0Improved
add distance-to-surface factor to ray coverage (ablation)camera placement · ScanReferAcc@0.25: 55.7 -> 59.2Improved
add distance-to-trajectory factor (ablation)camera placement · ScanReferAcc@0.25: 59.2 -> 61.0Improved
zero-initialized 1x1 conv in adapter vs Gaussian init (ablation)semantic adapter · ScanReferAcc@0.25: 59.5 -> 60.4Improved
use panoramic views instead of pinhole views (ablation)PanoGrounder · ScanReferAcc@0.25: 51.0 (pinhole 16 views) -> 61.0Improved
joint training on ScanRefer and ReferIt3DPanoGrounder · ScanRefer / Nr3Doverall accuracy: ScanRefer 61.0 -> 62.0; Nr3D 74.6 -> 76.1Improved
use mesh instead of 3DGS rendering; fully automated 3DGS reconstruction (ablation)PanoGrounder · ScanReferAcc@0.25: mesh 61.6; 3DGS 60.8; automated 3DGS* 55.1Mixed
scale VLM backbone from Qwen2.5-VL-3B to 7B to CogVLM-17B (ablation)PanoGrounder · ScanReferAcc@0.25: 45.6 -> 52.1 -> 61.0Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 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