AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

UZ3DVG: Unaided Zero-Shot 3D Visual Grounding with Generated Language Conditions

Wenbin Tan, Jiawen Lin, Yuan Xie, Yachao Zhang, Yanyun Qu

CVPR 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Adapting VLMs for zero-shot classification

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 TSVDE (text-guided sampling) over random sampling (ablation)UZ3DVG · ScanRefer valAcc@0.25: 40.75 -> 41.83Improved
add Reasoning Chain Distillation (RCD) (ablation)UZ3DVG with TSVDE and GeoSM · ScanRefer valAcc@0.25: 43.15 -> 45.42Improved
add GeoSM on top of TSVDE (ablation)UZ3DVG with TSVDE and RCD · ScanRefer valAcc@0.25: 43.26 -> 45.42Improved
all three modules combined vs none (ablation)UZ3DVG · ScanRefer valAcc@0.25: 40.75 -> 45.42Improved
generate pseudo-labels with Doubao-seed-1.6 instead of Qwen3-VL-Plus (ablation)UZ3DVG · ScanRefer valAcc@0.25: 43.87 -> 45.42Improved
increase pseudo-labels from 10K to 20K (ablation)UZ3DVG · ScanRefer valAcc@0.25: 25.71 -> 36.88Improved
increase pseudo-labels from 30K to 40K (ablation)UZ3DVG · ScanRefer valAcc@0.25: 43.36 -> 45.42Improved
project 2D boxes into 3D with MSAG-RC instead of 2D-only descriptions (ablation)UZ3DVG · ScanRefer valAcc@0.50: 22.83 -> 39.47Improved
add reasoning chains to MSAG-RC descriptions (ablation)UZ3DVG · ScanRefer valAcc@0.50: 39.47 -> 41.08Improved
use UZ3DVG instead of ZSVG3D without Mask3Dzero-shot 3DVG · ScanRefer val (non-Mask3D)Acc@0.25: 20.00 -> 43.13Improved
UZ3DVG vs SeeGround on Nr3D transferzero-shot 3DVG · Nr3D valaccuracy: 46.1 -> 46.5Improved
UZ3DVG vs SeeGround on Nr3D Dep. subsetzero-shot 3DVG · Nr3D val (View-Dependent)accuracy: 42.3 -> 39.3Hurt

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