AI Paper Atlas

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

Deformable Feature Aggregation for Dynamic Multi-modal 3D Object Detection

Zehui Chen, Zhenyu Li, Shiquan Zhang, Liangji Fang, Qinhong Jiang, Feng Zhao

ECCV 2022 · Detection and segmentation › 3D detection and driving · approach: Multi-camera BEV detection · 149 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 Cross-Domain DeformCAFA with image branch (ablation)CenterPoint · nuScenes val 1/8mAP: 50.28 -> 56.96Improved
add image-level dropout training (ablation)CenterPoint + DeformCAFA · nuScenes val 1/8mAP: 56.96 -> 57.03No clear effect
add depth-aware GT-AUG (ablation)CenterPoint + DeformCAFA + dropout · nuScenes val 1/8mAP: 57.03 -> 58.45Improved
use multiply of point and image features for query token instead of concat (ablation)Cross-Domain DeformCAFA · nuScenes val 1/8mAP: 58.01 -> 58.45Improved
use both point and image features for token instead of points only (ablation)Cross-Domain DeformCAFA · nuScenes val 1/8mAP: 57.10 -> 58.45Improved
replace AutoAlign fusion with DeformCAFA (ablation)CenterPoint · nuScenes val 1/8mAP: 56.69 -> 58.45Improved
use depth-aware GT-AUG instead of MoCa GT-AUG (ablation)CenterPoint · nuScenes val 1/8mAP: 53.08 -> 53.48Improved
mix-up ratio alpha 1.0 instead of 0.5 to 0.8 (ablation)depth-aware GT-AUG · nuScenes val 1/8mAP: drops about 0.7 mAPHurt
train with 3 images instead of 1 (ablation)image-level dropout · nuScenes val 1/8mAP: 57.93 -> 58.45Improved
train with 6 images instead of 3 (ablation)image-level dropout · nuScenes val 1/8mAP: 58.45 -> 58.51No clear effect
infer with 3 images instead of 6AutoAlignV2 · nuScenes val 1/8mAP: 58.45 -> 54.32Hurt
infer with 0 images (LiDAR only)AutoAlignV2 · nuScenes val 1/8mAP: 58.45 -> 50.29Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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