AI Paper Atlas

Research map / 3D vision / 3D generation

3D Aware Region Prompted Vision Language Model

An-Chieh Cheng, Yang Fu, Yukang Chen, Zhijian Liu, Xiaolong Li, Subhashree Radhakrishnan, Song Han, Yao Lu, Jan Kautz, Pavlo Molchanov, Hongxu Yin, Xiaolong Wang and 1 more

ICLR 2026 · 3D vision › 3D generation · approach: Semantic scene completion · 27 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 single-view 2D pre-training before multi-view fine-tuning (ablation)SR-3D (3D PE) · ScanQA valCIDEr: 102.9 -> 109.3Improved
add single-view 2D pre-training before multi-view fine-tuning (ablation)SR-3D (3D PE) · SQA3D testexact match: 59.1 -> 62.2Improved
add 3D positional embeddings without 2D pre-training (ablation)SR-3D 3D fine-tuned · ScanQA valCIDEr: 101.3 -> 102.9No clear effect
add 3D positional embeddings on top of 2D pre-training (ablation)SR-3D 3D fine-tuned · ScanQA valCIDEr: 108.2 -> 109.3Improved
add 2D pre-training (ablation)SR-3D · SR-3D-Bench global Distancesuccess rate: 60.3 -> 74.8Improved
add 2D pre-training (ablation)SR-3D · SR-3D-Bench Multi. Complexaccuracy: 59.0 -> 76.0Improved
add 3D positional embeddings and region extractor to NVILA-Lite-8BSR-3D-8B · RealWorldQAaccuracy: 65.6 -> 68.1Improved
add 3D positional embeddings and region extractor to NVILA-Lite-8BSR-3D-8B · OCR benchmarks (TextVQA, ChartQA, DocVQA)accuracy: TextVQA 78.1 -> 77.3; DocVQA 91.7 -> 91.0Hurt
add 3D positional embeddings and region extractor to NVILA-Lite-8BSR-3D-8B · General knowledge (GQA, AI2D, POPE)accuracy: GQA 65.3 -> 64.2No clear effect
replace ground-truth point clouds with Cut3R-estimated point maps at test timeSR-3D · ScanQA valCIDEr: 109.3 -> 109.3No clear effect
replace ground-truth point clouds with Cut3R-estimated point maps at test timeVideo-3D LLM · ScanQA valCIDEr: 102.1 -> 100.7Hurt
add region prompts and 3D positional embeddings (versus SoM-prompted VLMs)SR-3D-8B · SR-3D-Bench qualitativeaccuracy: 47.1 -> 79.5Improved

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