AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

3D-Aware VLMs with Implicit and Explicit Geometries

Wenhao Li, Xueying Jiang, Quanhao Qian, Deli Zhao, Ran Xu, Shijian Lu, Gongjie Zhang

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 1 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 implicit geometry tokens (IGTs) (ablation)Qwen2.5-VL-3B · 3D video detectionF1_25: 30.9 -> 40.5Improved
add explicit geometry tokens (EGTs) (ablation)Qwen2.5-VL-3B · 3D video detectionF1_25: 30.9 -> 34.7Improved
add EGTs on top of IGTs (ablation)VLM-IE3D · 3D video detectionF1_25: 40.5 -> 42.8Improved
fuse IGTs and EGTs with implicit-explicit attention instead of concat (ablation)VLM-IE3D · 3D video detectionF1_25: 41.5 -> 42.8Improved
fuse with IEA instead of simple addition (ablation)VLM-IE3D · 3D video detectionF1_25: 42.4 -> 42.8Improved
use learnable weighted fusion instead of IEA (ablation)VLM-IE3D · 3D video detectionF1_25: 42.8 -> 41.2Hurt
use depth, point or Gaussian attribute for EGTs (ablation)VLM-IE3D · 3D video detectionF1_25: 42.5 to 42.8No clear effect
use deep encoder (DepthAnything V2) for explicit embedding instead of light patch embedding (ablation)VLM-IE3D · 3D video detectionF1_25: 42.8 -> 35.9Hurt
use average pooling instead of one-layer patch embedding for EGTs (ablation)VLM-IE3D · 3D video detectionF1_25: 42.8 -> 42.0Hurt
swap AnySplat for VGGT or pi^3 as geometry encoder (ablation)VLM-IE3D · 3D video detectionF1_25: 42.8 -> 41.7 (VGGT), 42.1 (pi^3)Mixed
add implicit and explicit geometry to VLMQwen2.5-VL-3B · Scan2CapC@0.5: 58.0 -> 80.4Improved
add EGTs on top of implicit-only VG LLMVG LLM · ScanRefer Acc@0.50Acc@0.50: 11.8 -> 16.9Improved

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