AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

Agentic Collaborative Cognition for Zero-Shot 3D Understanding

Wenxin Wang, Bo Zhang, Feng Chen, Zixuan Wang, Wen Li, Changsheng Li, Yinjie Lei

ECCV 2026 · Vision-language and multimodal › Robot action and navigation · approach: Pretraining and data augmentation for VLN · 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 object information table to the BEV cognitive map (ablation)Ours · Nr3DAcc@0.25: -8.9% Acc@0.25 when removedImproved
add object information table to the BEV cognitive map (ablation)Ours · SQA3DEM: -3.9% EM when removedImproved
add candidate filtering and feedback (ablation)Ours · Nr3DAcc@0.25: -7.4% Acc@0.25 when disabledImproved
add rendered images to real images (ablation)Ours · ScanQAEM: -5.1% EM without rendered imagesImproved
add real images to rendered images (ablation)Ours · ScanQAEM: -4.7% EM without real imagesImproved
use cognitive-map viewpoint planning instead of random viewpoints (ablation)Ours · Nr3DAcc@0.25: +42.9% Acc@0.25Improved
use cognitive-map viewpoint planning instead of co-visible template (ablation)Ours · Nr3DAcc@0.25: +13.6% Acc@0.25Improved
increase Perception Agents from 1 to 3 (ablation)Ours · Nr3D (250 queries)Acc@0.25: 65.6 -> 66.8Improved
increase planned viewpoints K from 8 to 9 (ablation)Ours · Nr3D, ScanRefer, SQA3DaccuracyHurt
raise viewpoint matching threshold above 0.8 (ablation)Ours · multiple benchmarksaccuracyHurt
replace Qwen2.5-VL-72B with GPT-4o (ablation)Ours · Nr3DAcc@0.25: 56.7 -> 61.2Improved
shrink MLLM from Qwen2.5-VL-72B to 7B (ablation)Ours · Nr3DAcc@0.25: 56.7 -> 49.3Hurt

Earlier methods this paper compares with

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