AI Paper Atlas

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

Dense Multimodal Alignment for Open-Vocabulary 3D Scene Understanding

Ruihuang Li, Zhengqiang ZHANG, Chenhang He, Zhiyuan Ma, Vishal Patel, Lei Zhang

ECCV 2024 · Vision-language and multimodal › CLIP and zero-shot · approach: Open-vocabulary alignment from VLMs · 6 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
align 3D features with generated text modality in addition to 2D featuresDMA (OpenSeg)-3D · ScanNet valmIoU: 47.9 -> 53.3Improved
use mutually inclusive BCE loss instead of cross-entropy for text alignment (ablation)DMA · ScanNetmIoU: 46.9 -> 53.3Improved
add MLLM scene descriptions to tagging-model text (ablation)DMA · ScanNetmIoU: 46.3 -> 50.5Improved
add MLLM descriptions to tagging text (ablation)DMA · nuScenesmIoU: 43.3 -> 43.8No clear effect
combine frozen CLIP features with learnable mask features (dual-path) (ablation)DMA · ScanNetmIoU: 42.0 -> 44.8Improved
fine-tune the FC-CLIP mask head instead of freezing it (ablation)DMA · ScanNetmIoU: 40.1 -> 42.0Improved
full mask-head fine-tuning versus LoRA or Adapter (ablation)DMA · ScanNetmIoU: 42.0 vs 41.3 (LoRA) vs 40.9 (Adapter)Improved
add GPT denoising of RAM tags (ablation)DMA · ScanNetmIoU: +2.6 (basic), +1.3 (enhanced)Improved
use FC-CLIP instead of OpenSeg as 2D encoderDMA · ScanNet valmIoU: 53.3 -> 51.8Hurt
use FC-CLIP instead of OpenSeg as 2D encoderDMA · nuScenes valmIoU: 45.1 -> 47.4Improved
DMA versus fully-supervised model on long-tail ScanNet200 classesDMA (FC-CLIP)-3D · ScanNet200 tailmIoU: 3.4 -> 5.8Improved

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