AI Paper Atlas

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

BEVHeight++: Toward Robust Visual Centric 3D Object Detection

Lei Yang, Tao Tang, Jun Li, Kun Yuan, Kai Fa Wu, Peng Chen, Li Wang, Yi Huang, Lei Li, Xinyu Zhang, Kaicheng Yu

TPAMI 2025 · Detection and segmentation › 3D detection and driving · approach: Multi-camera BEV detection · 85 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
use dynamic-increasing discretization instead of uniform (ablation)BEVHeight · DAIR-V2X-I val vehicle MidAP3D: 63.75 -> 64.71Improved
set DID alpha to 2.0 instead of 1.5 (ablation)BEVHeight · DAIR-V2X-I val vehicle MidAP3D: 64.54 -> 64.71Improved
replace depth-based view transformation with height-based (ablation)BEVDepth · DAIR-V2X-I val vehicle MidAP3D: 63.58 -> 65.77Improved
replace depth-based view transformation with height-based (ablation)BEVDet · DAIR-V2X-I val vehicle MidAP3D: 51.92 -> 60.48Improved
add LiDAR point cloud supervision (ablation)BEVDepth · DAIR-V2X-I val vehicle MidAP3D: 60.75 -> 60.37Hurt
add LiDAR point cloud supervision (ablation)BEVHeight · DAIR-V2X-I val vehicle MidAP3D: 63.49 -> 63.61No clear effect
add image-view fusion (height distribution concatenated) (ablation)BEVHeight++ · nuScenes valNDS: 0.475 -> 0.492Improved
add BEV fusion via deformable cross-attention (ablation)BEVHeight++ · nuScenes valNDS: 0.475 -> 0.489Improved
add both image-view and BEV fusion (ablation)BEVHeight++ · nuScenes valNDS: 0.475 -> 0.498Improved
add depth distribution to image-view fusion features (ablation)BEVHeight++ · nuScenes valNDS: 0.498 -> 0.476Hurt
add height distribution to image features in fusion (ablation)BEVHeight++ · nuScenes valNDS: 0.489 -> 0.498Improved
use two-stage training (depth branch first, then others) (ablation)BEVHeight++ · nuScenes valmAP: 0.364 -> 0.373Improved

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