AI Paper Atlas

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

Domain Generalization of 3D Object Detection by Density-Resampling

Shuangzhi Li, Lei Ma, Xingyu Li

ECCV 2024 · Detection and segmentation › 3D detection and driving · approach: Unified LiDAR 3D detection · 5 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 PDDA density-resampling augmentation during source training (ablation)VoxelRCNN source-only · NuScenes -> KITTImAP BEV/3D: 35.34/18.5 -> 41.07/24.08Improved
add auxiliary self-supervised scene restoration (multi-task learning) (ablation)VoxelRCNN + PDDA · NuScenes -> KITTImAP BEV/3D: 41.07/24.08 -> 41.35/24.12No clear effect
add test-time adaptation using the restoration task (ablation)VoxelRCNN + PDDA + MTL · NuScenes -> KITTImAP BEV/3D: 41.35/24.12 -> 42.17/25.15Improved
add test-time adaptation using the restoration task (ablation)VoxelRCNN + PDDA + MTL · Waymo -> NuScenesmAP BEV/3D: 17.09/11.17 -> 17.22/11.26Improved
replace density-downsampling test-time training with KNN-masking TTT (ablation)test-time adaptation · NuScenes -> KITTIFPS: 4.23 -> 0.52Hurt
add perceptual loss to MSE restoration loss (ablation)test-time adaptation · NuScenes -> KITTImAP BEV/3D and FPS: 42.17/25.15 -> 42.21/25.16; FPS 4.23 -> 0.89No clear effect
increase test-time update iterations beyond optimum (ablation)test-time adaptation · NuScenes -> KITTI and Waymo -> NuScenesmAPHurt
reduce test-time iterations to N_iter=1 for real-time speed (ablation)test-time adaptation · NuScenes -> KITTImAP BEV/3D: 42.17/25.15 -> 41.41/24.19No clear effect
apply Ours versus source-only on domains without density gapOurs · Waymo -> KITTImAP BEV/3D: 65.41/46.79 -> 65.47/46.67No clear effect
use PDDA instead of UDA augmentation RBRS that needs target info (ablation)source training augmentation · Waymo -> NuScenesmAP BEV/3D: 17.33/11.20 -> 16.83/11.04Mixed

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