AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

GAP-MLLM: Geometry-Aligned Pre-training for Activating 3D Spatial Perception in Multimodal Large Language Models

Jiaxin Zhang, Junjun Jiang, Haijie LI, Youyu Chen, Kui Jiang, Dave Zhenyu Chen

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 3 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 sparse joint geometry-semantics pre-training (ablation)GAP-MLLM-3B · 3D video detectionF1: 44.7 -> 48.7Improved
add multi-level gated fusion (ablation)GAP-MLLM-3B · 3D video detectionF1: 44.7 -> 47.5Improved
add both pre-training and gated fusion (ablation)GAP-MLLM-3B · 3D video detectionF1: 44.7 -> 50.6Improved
add GAP to VG-LLM (ablation)VG-LLM-4B · 3D video detectionF1: 38.2 -> 43.5Improved
add joint pre-training to pure Qwen3-VL (ablation)Qwen3-VL · 3D video detectionF1: 39.7 -> 42.4Improved
add semantic labels to pointmap-only pre-training (ablation)VG-LLM · 3D video detectionF1: 40.1 -> 41.4Improved
add pointmap pre-training (ablation)VG-LLM · 3D video detectionF1: 38.2 -> 40.1Improved
add two-stage grounding (ablation)VG-LLM · ScanReferAcc@0.25: 37.5 -> 45.4Improved
add sparse joint pre-training (ablation)GAP-MLLM-3B · ScanReferAcc@0.25: 47.9 -> 51.7Improved
use gated fusion over addition, weighted fusion and cross-attention (ablation)GAP-MLLM · fusion strategy ablationaccuracyImproved
blur first frame (sigma=12) (ablation)GAP-MLLM · 3D video detectionF1: 50.6 -> 46.0Hurt
occlude 30% of first frame (ablation)GAP-MLLM · 3D video detectionF1: 50.6 -> 50.1No clear effect

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