AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

ROSE: Real-Time Open-World Scene Understanding from Monocular Video via Compact Multimodal 4D Scene Graphs

Ziyue Qiu, Yong Wang, Jin Pan

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 4DSG in prompt at inference (training-free)DSR Qwen2.5-VL-7B · DSR-Bench (accuracy, 13 subtasks avg)accuracy: 58.9 -> 62.3Improved
add 4DSG as context during fine-tuningROSE (Training-free) · DSR-Bench (accuracy, 13 subtasks avg)accuracy: 62.3 -> 67.5Improved
add token-matched null context (LenCtrl)DSR · DSR-Bench (accuracy, 13 subtasks avg)accuracy: 58.9 -> 59.5No clear effect
add 4DSG fine-tuningDSR · DSR-Bench held-out subtask familyaccuracy: 22.9 -> 45.4Improved
add 4DSG fine-tuningDSR · DSR-Bench Absolute Directionaccuracy: 73.8 -> 83.7Improved
add 4DSG fine-tuningDSR · DSR-Bench Direction Predictionaccuracy: 35.5 -> 38.6Improved
add 2D anchors to raw video (ablation)Qwen2.5-VL-32B · VLM4D balanced 200-question subset (Qwen2.5-VL-32B)overall accuracy: 49.5 -> 56.0Improved
add 3D geometry and trajectories to anchors (ablation)Qwen2.5-VL-32B · VLM4D balanced 200-question subset (Qwen2.5-VL-32B)overall accuracy: 56.0 -> 64.5Improved
use bounding-box context crops instead of masked crops (ablation)Qwen2.5-VL-32B 4DSG · VLM4D balanced 200-question subset (Qwen2.5-VL-32B)overall accuracy: 64.5 -> 73.5Improved
use bounding-box context crops instead of masked crops (ablation)Qwen2.5-VL-32B 4DSG · VLM4D FPaccuracy: 91.1 -> 91.1No clear effect
append 4DSG to promptQwen2.5-VL-32B · DSR-Bench prompt latencyprompt latency (s): 3.95 -> 12.45Hurt
use ROSE (monocular 4DSG) instead of SNOW (RGB plus point clouds)4DSG pipeline · VLM4D Direct.accuracy: 71.8 -> 79.8Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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