AI Paper Atlas

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

DART: Domain-Agnostic Residual Transfer for Generalist Anomaly Detection

Muhammad Aqeel, Maham Nazir, Marco Cristani, Francesco Setti

NeurIPS 2026 · Label-efficient and robust learning › OOD and anomaly detection · approach: Unified multi-class anomaly detection · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes DART, a domain-agnostic residual transfer approach for generalist anomaly detection that separates transferable anomaly cues from domain-specific appearance residuals using normal-abnormal reference samples.

Problem

Generalist anomaly detection must detect anomalies in unseen domains from only a few reference samples; existing normal-abnormal guided residuals mix transferable anomaly cues with domain-specific appearance factors (texture, geometry, acquisition conditions), limiting cross-domain transfer.

Approach

Targets residual domain entanglement in normal-abnormal guided GAD by making residuals domain-agnostic so that anomaly cues transfer across domains rather than carrying domain-specific appearance factors. Full method details are not available from the abstract.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add DART residual disentanglement to NAGLNAGL · VisA N^2+A^1image-level AUROC: 87.90 -> 89.72Improved
naively add abnormal reference without disentanglement (+Inv)ResAD · VisA N^1image-level AUROC: -6.5 I-AUROCHurt
add DART on top of naive abnormal integrationInCTRL+Inv · VisA N^1image-level AUROC: 79.87 -> 86.06Improved
add DART to InCTRL under industrial-to-medical shiftInCTRL · MVTec AD -> BraTS N^1image-level AUROC: 65.20 -> 78.48Improved
add DART to NAGL at N^1 on BraTSNAGL · MVTec AD -> BraTS N^1pixel-level AUROC: 93.00 -> 90.33Hurt
add sufficiency loss L_suf (ablation)DART (InCTRL) · MVTec AD k=2image-level AUROC: 93.57 -> 96.21Improved
add adversarial invariance loss L_adv (ablation)DART (InCTRL) · MVTec AD k=2image-level AUROC: 95.57 -> 96.21Improved
add reconstruction loss L_recon (ablation)DART (InCTRL) · MVTec AD k=2image-level AUROC: 95.48 -> 96.21Improved
add cross-domain meta-training on top of RDM (ablation)DART (InCTRL) · MVTec AD k=2image-level AUROC: 95.54 -> 96.21Improved
add CDMT without RDM (ablation)InCTRL · MVTec AD and VisA k=2image-level AUROC: +0.00No clear effect
use transformer encoder instead of linear projection (ablation)DART encoder (InCTRL) · VisA k=2image-level AUROC: 81.29 -> 88.64Improved
add LayerNorm in anomaly encoder (ablation)DART encoder (InCTRL) · VisA k=2image-level AUROC: 84.65 -> 88.64Improved

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