AI Paper Atlas

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

Sequential Multi-View Fusion Network for Fast LiDAR Point Motion Estimation

Gang Zhang, Xiaoyan Li, Zhenhua Wang

ECCV 2022 · Detection and segmentation › 3D detection and driving · approach: LiDAR perception for driving · 4 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 BEV branch with multi-scan input to RV-only model (ablation)SMVF · SemanticKITTI MOS valmoving IoU: 63.8 -> 73.5Improved
add single-scan RV branch to BEV-only model (ablation)SMVF · SemanticKITTI MOS valmoving IoU: 73.5 -> 76.9Improved
feed multiple scans instead of a single scan to the RV branch (ablation)SMVF · SemanticKITTI MOS valmoving IoU: 76.9 -> 76.7No clear effect
increase history scans from k=1 to k=2 (ablation)SMVF · SemanticKITTI MOS valmoving IoU: 75.1 -> 76.9Improved
increase history scans from k=2 to k=4 (ablation)SMVF · SemanticKITTI MOS valmoving IoU: 76.9 -> 76.9No clear effect
add BEV branch to RV-only model (ablation)SMVF · Waymo val Vehiclemean L2 error (m/s): 0.70 -> 0.17Improved
use multiple scans in RV branch instead of single scan (ablation)SMVF · Waymo val Vehiclemean L2 error (m/s): 0.17 -> 0.17Hurt
add SICP augmentationSMVF · Waymo val Cyclistmean L2 error (m/s): 0.38 -> 0.35Improved
add SICP augmentationMotionNet* · Waymo val Cyclistmean L2 error (m/s): 0.35 -> 0.31Improved
add SICP when pedestrian category treated as stationarySMVF · Waymo generalizationpedestrian mean error (m/s): 0.79 -> 0.70Improved
add SICP when pedestrian category is ignoredSMVF · Waymo generalizationpedestrian mean error (m/s): 0.57 -> 0.36Improved
label ablated cyclist as ignored instead of stationary (ablation)FastFlow3D · Waymo generalizationcyclist mean error (m/s): 1.13 -> 0.83Improved

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