AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

If It's Not Efficient, It's Not Usable: Real-Time OOD Detection with Latent De-Biasing and High-Quality Negative Samples

Yuchuan Li, Jae-Mo Kang, Il-Min Kim

ECCV 2026 · Label-efficient and robust learning › OOD and anomaly detection · approach: Spurious-aware OOD detection · 0 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 latent-space pixel-intensity de-biasing (32 perturbed variants) (ablation)VAE OOD detector, no negatives · CIFAR-100 ID, Tiny-ImageNetAUROC: 47.5 -> 58.0Improved
add latent-space pixel-intensity de-biasing (32 perturbed variants) (ablation)VAE OOD detector, no negatives · ImageNet-1k ID, NINCOAUROC: 43.5 -> 57.0Improved
increase perturbed variants from 4 to 32 (ablation)VAE OOD detector · CIFAR-100 ID, SVHNAUROC: 56.9 -> 63.7Improved
add Mosaic-based initialization of negatives (step 1) (ablation)Base VAE · CIFAR-100 ID, SVHNAUROC: 63.7 -> 67.6Improved
add iterative negative update (step 2) (ablation)Base + s1 · CIFAR-100 ID, SVHNAUROC: 67.6 -> 71.4Improved
add screening of easy and invalid negatives (step 3) (ablation)Base + s1 + s2 · CIFAR-100 ID, SVHNAUROC: 71.4 -> 75.5Improved
add screening of negatives (step 3) (ablation)Base + s1 + s2 · ImageNet-1k ID, NINCOAUROC: 60.2 -> 62.0Improved
use proposed negative generation instead of HamOS (ablation)Stage 2 replacement · CIFAR-100 ID, Tiny-ImageNetAUROC: 66.0 -> 67.3Improved
use proposed negative generation instead of CSI (ablation)Stage 2 replacement · ImageNet-1k ID, NINCOAUROC: 53.4 -> 62.0Improved
use proposed method instead of strongest VAE baseline DPTinyML latency-matched OOD detectors · CIFAR-100 ID, Tiny-ImageNetAUROC: 65.9 -> 67.3Improved
replace importance-sampling de-biasing with latent de-biasingVAE OOD detector · ImageNet-1k, Jetson latencylatency (ms): 1.27 -> 0.12 msImproved

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