AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Pixel-Wise Energy-Biased Abstention Learning for Anomaly Segmentation on Complex Urban Driving Scenes

Yu Tian, Yuyuan Liu, Guansong Pang, Fengbei Liu, Yuanhong Chen, Gustavo Carneiro

ECCV 2022 · Detection and segmentation › Open-set and few-shot detection · approach: Industrial anomaly detection · 107 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 PAL loss with energy-biased penalty to EBM (ablation)PEBAL · LostAndFound test (WideResnet38)AP: 70.24 -> 77.19Improved
add PAL loss with energy-biased penalty to EBM (ablation)PEBAL · LostAndFound test (WideResnet38)FPR95: 8.92 -> 1.19Improved
add EBM with adaptive penalty to fixed-penalty PAL (ablation)PEBAL · LostAndFound test (WideResnet38)AP: 72.73 -> 77.19Improved
replace entropy-maximisation baseline with joint EBM + PAL (ablation)PEBAL · LostAndFound test (WideResnet38)AP: 69.02 -> 77.19Improved
add smoothness and sparsity regularisation (ablation)PEBAL · LostAndFound test (WideResnet38)AP: 77.19 -> 78.29Improved
add smoothness and sparsity regularisation (ablation)PEBAL · LostAndFound test (WideResnet38)FPR95: 1.19 -> 0.81Improved
increase share of COCO outlier classes from 1% to 75% (ablation)PEBAL · Fishyscapes LostAndFound valAP: 53.57 -> 57.86No clear effect
increase share of COCO outlier classes from 1% to 75% (ablation)PEBAL · Fishyscapes Static valAP: 85.84 -> 91.85Improved
increase outlier training set size from 5% to 50% (ablation)PEBAL · Fishyscapes LostAndFound valAP: 54.32 -> 57.34No clear effect
increase outlier training set size from 5% to 50% (ablation)PEBAL · Fishyscapes Static valAP: 89.11 -> 91.29Improved
set lambda to 0.01 instead of 0.1 (ablation)PEBAL · LostAndFound test (WideResnet38)AP: 77.15 -> 78.29No clear effect
fine-tune only the final classification block instead of re-training whole networkPEBAL · trainable parametersparameters: 137.1M -> 1.3MImproved

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