AI Paper Atlas

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

FlowCon: Out-of-Distribution Detection using Flow-based Contrastive Learning

Saandeep Aathreya, Shaun Canavan

ECCV 2024 · Label-efficient and robust learning › OOD and anomaly detection · approach: Spurious-aware OOD detection · 1 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
train a flow model with Bhattacharyya-based flow contrastive loss on fixed ResNet18 penultimate featuresFlowCon · CIFAR-10 far-OOD (ResNet18)AUROC: 96.47 (Heatmap) -> 97.19Improved
FlowCon versus post-hoc methods in mixed near/far OODFlowCon · CIFAR-10 vs CIFAR-100 (ResNet18)AUROC: 86.45 (MSP) -> 93.97Improved
FlowCon versus Energy in near-OOD with WideResNetFlowCon · CIFAR-100 vs CIFAR-10 (WideResNet)AUROC: 74.30 (Energy) -> 67.03Hurt
lower feature dimension (WideResNet, 128-d) instead of ResNet18 (512-d)FlowCon · CIFAR-100 near-OODAUROC: 82.80 -> 67.03Hurt
use flow loss weight lambda 0.07 instead of larger or smaller values (ablation)FlowCon · CIFAR-100 far-OOD (WideResNet)AUROC: 75.62 (0.05), 78.57 (1.0) -> 83.62 (0.07)Improved
classify with class-wise learned distributions (Bayes rule) instead of original classifier headFlowCon · CIFAR-100 (ResNet18)accuracy: 75.8 -> 74.9No clear effect
classify with FlowCon distributions on WideResNetFlowCon · CIFAR-10 (WideResNet)accuracy: 93.3 -> 93.8Improved
add contrastive loss to flow likelihood (class-specific multimodal latent) (ablation)FlowCon · toy moonslikelihood separationImproved

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