AI Paper Atlas

Research map / Detection and segmentation / CNN features

Self-Distillation for Robust LiDAR Semantic Segmentation in Autonomous Driving

Jiale Li, Hang Dai, Yong Ding

ECCV 2022 · Detection and segmentation › CNN features · approach: Lane detection · 35 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
replace average VFE with TransVFE (ablation)U-Net baseline · SemanticKITTI valmIoU: 64.90 -> 65.47Improved
replace average VFE with TransVFE (ablation)U-Net baseline · nuScenes valmIoU: 75.94 -> 76.42Improved
replace PointNet VFE with TransVFE (ablation)U-Net baseline · SemanticKITTI valmIoU: 65.08 -> 65.47Improved
add compound-transformation TTA instead of no TTA (ablation)TransVFE baseline · SemanticKITTI valmIoU: 65.47 -> 67.97Improved
use compound transform instead of four individual transformations in TTA (ablation)TTA, M=5 · SemanticKITTI valmIoU: 66.64 -> 67.75Improved
use compound transform instead of four individual transformations in TTA (ablation)TTA, M=5 · nuScenes valmIoU: 77.27 -> 77.54Improved
increase TTA variants beyond M=6 (ablation)compound TTA · SemanticKITTI valmIoU: 67.97 -> 68.04No clear effect
add TTA to teacher soft labels with EMA teacher (ablation)self-distillation · SemanticKITTI valmIoU: 66.18 -> 66.64Improved
add TTA to teacher soft labels with fixed pre-trained teacher (ablation)self-distillation · SemanticKITTI valmIoU: 65.89 -> 66.09Improved
update teacher by EMA of student instead of fixed pre-trained teacher (ablation)self-distillation with TTA · SemanticKITTI valmIoU: 66.09 -> 66.64Improved
initialize student and teacher from pre-trained baseline (ablation)self-distillation EMA+TTA · SemanticKITTI valmIoU: 66.64 -> 67.11Improved
add self-distillation (Exp. H) instead of baseline (ablation)TransVFE baseline · nuScenes valmIoU: 76.42 -> 77.74Improved

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