AI Paper Atlas

Research map / 3D vision / 3D generation

UniFusion: Unified Multi-View Fusion Transformer for Spatial-Temporal Representation in Bird's-Eye-View

Zequn Qin, Jingyu Chen, Chao Chen, Xiaozhi Chen, Xi Li

ICCV 2023 · 3D vision › 3D generation · approach: 3D perception for autonomous driving · 68 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 temporal fusion steps (0 to 1) (ablation)UniFusion · nuScenes 100m x 100mRoad mIoU: 79.04 -> 79.48Improved
infer with 6 temporal steps while training with only 1 step (ablation)UniFusion · nuScenes 100m x 100mRoad mIoU: 79.48 -> 81.12Improved
train with more fusion steps (1 vs 4) given 6 inference steps (ablation)UniFusion · nuScenes 100m x 100mRoad mIoU: 81.12 -> 81.25No clear effect
extend temporal fusion up to 10 steps in unified fusion versus warp-based BEVFormer (ablation)UniFusion · nuScenesmIoUImproved
add self-regression mechanism (12 layers) (ablation)UniFusion · nuScenes 100m x 100mRoad mIoU: 81.13 -> 81.97Improved
add self-regression mechanism (6 layers) (ablation)UniFusion · nuScenes 100m x 100mRoad mIoU: 80.42 -> 80.91Improved
increase Transformer layers from 6 to 12 (ablation)UniFusion · nuScenes 100m x 100mLane mIoU: 24.99 -> 25.76Improved
use learned adaptive temporal fusion instead of equal-weighted averaging (6 steps) (ablation)UniFusion · nuScenes 100m x 100mLane mIoU: 25.08 -> 25.76Improved
UniFusion ResNet50 versus BEVFormer with larger backbone and more queriesUniFusion · nuScenes 100m x 100mRoad mIoU: 80.1 -> 82.0Improved
switch from vanilla split to city-based splitUniFusion and baselines · nuScenes 160m x 100m easyAll mIoU: 40.9 -> 16.7Hurt
larger VoVNet99 backbone instead of Swin-tinyUniFusion · nuScenes 60m x 30mAll mIoU: 53.6 -> 57.4 (vanilla); 26.5 -> 25.6 (city-based)Mixed

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