AI Paper Atlas

Research map / Detection and segmentation / 3D detection and driving

Bridging the Domain Gap: Self-Supervised 3D Scene Understanding with Foundation Models

Zhimin Chen, Longlong Jing, Yingwei Li, Bing Li

NeurIPS 2023 · Detection and segmentation › 3D detection and driving · approach: 3D objectness and scene segmentation · 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
add semantic guided (foreground-aware) reconstruction to Point-MAE (ablation)Bridge3D · ScanNetV2 detectionAP25: 62.3 -> 63.2Improved
add semantic guided reconstruction to Point-MAE (ablation)Bridge3D · S3DIS segmentationmIoU: 64.8 -> 66.2Improved
add scene-level distillation (ablation)Bridge3D · ScanNetV2 detectionAP50: 41.1 -> 43.0Improved
add scene-level distillation (ablation)Bridge3D · S3DIS segmentationmIoU: 66.2 -> 68.4Improved
add object-level distillation (ablation)Bridge3D · ScanNetV2 detectionAP50: 43.0 -> 44.2Improved
add object-level distillation (ablation)Bridge3D · S3DIS segmentationmAcc: 73.7 -> 76.1Improved
use text modality only instead of none (ablation)Bridge3D · S3DIS segmentationmIoU: 61.1 -> 67.8Improved
use image modality only instead of none (ablation)Bridge3D · S3DIS segmentationmIoU: 61.1 -> 68.3Improved
use both text and image instead of text only (ablation)Bridge3D · ScanNetV2 detectionAP50: 42.5 -> 44.2Improved
use both text and image instead of image only (ablation)Bridge3D · ScanNetV2 detectionAP50: 43.3 -> 44.2Improved
replace MaskCLIP + superpixel masks with foundation-model masks (ablation)Bridge3D · ScanNetV2 detectionAP50: 43.0 -> 44.2Improved
replace MaskCLIP + superpixel masks with foundation-model masks (ablation)Bridge3D · S3DIS segmentationmIoU: 67.1 -> 70.2Improved

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