GATE-AD: Graph Attention Network Encoding For Few-Shot Industrial Visual Anomaly Detection

Aggelos Psiris; Yannis Panagakis; Maria Vakalopoulou; Georgios Th. Papadopoulos

cs.CVMar 16, 2026

GATE-AD: Graph Attention Network Encoding For Few-Shot Industrial Visual Anomaly Detection

Authors: Aggelos Psiris, Yannis Panagakis, Maria Vakalopoulou, Georgios Th. Papadopoulos

Organizations: Harokopio University of Athens, Athens, Greece · National and Kapodistrian University of Athens, Athens, Greece · Archimedes, Athena Research Center, Athens, Greece · CentraleSupelec, University Paris Saclay, Paris, France

Abstract

Few-Shot Industrial Visual Anomaly Detection (FS-IVAD) comprises a critical task in modern manufacturing settings, where automated product inspection systems need to identify rare defects using only a handful of normal/defect-free training samples. In this context, the current study introduces a novel reconstruction-based approach termed GATE-AD. In particular, the proposed framework relies on the employment of a masked, representation-aligned Graph Attention Network (GAT) encoding scheme to learn robust appearance patterns of normal samples. By leveraging dense, patch-level, visual feature tokens as graph nodes, the model employs stacked self-attentional layers to adaptively encode complex, irregular, non-Euclidean, local relations. The graph is enhanced with a representation alignment component grounded on a learnable, latent space, where high reconstruction residual areas (i.e., defects) are assessed using a Scaled Cosine Error (SCE) objective function. Extensive comparative evaluation on the MVTec AD, VisA, and MPDD industrial defect detection benchmarks demonstrates that GATE-AD achieves state-of-the-art performance across the $1$ - to $8$ -shot settings, combining the highest detection accuracy (increase up to $1.8\%$ in image AUROC in the 8-shot case in MPDD) with the lowest per-image inference latency (at least $25.05\%$ faster), compared to the best-performing literature methods. In order to facilitate reproducibility and further research, the source code of GATE-AD is available at https://github.com/gthpapadopoulos/GATE-AD.

Figures & tables

Figure 1: General architecture of the proposed masked, representation-aligned GAT encoding scheme for FS-IVAD.

Method

MVTec AD

VisA

MPDD

Image

Pixel

Image

Pixel

Image

Pixel

AUROC

mAP

AUROC

PRO

AUROC

mAP

AUROC

PRO

AUROC

mAP

AUROC

PRO

1-shot

SPADE [ 4 ] (arXiv’20)

82.9

\pm

2.6

91.7

\pm

1.2

92.0

\pm

0.3

85.7

\pm

0.7

80.7

\pm

5.0

82.3

\pm

4.3

96.2

\pm

0.4

85.7

\pm

1.1

PatchCore [ 23 ] (CVPR’22)

83.4

\pm

3.0

92.2

\pm

1.5

92.0

\pm

1.0

79.7

\pm

2.0

79.9

\pm

2.9

82.8

\pm

2.3

95.4

\pm

0.6

80.5

\pm

2.5

GraphCore [ 30 ] (ICLR’23)

89.9

95.6

84.7

95.2

Table 1: FS-IVAD detection results on the MVTec AD, VisA, and MPDD benchmarks. Methods with an asterisk (*) use the standardized sample selection process (Seed:

0

Figure 2: Indicative $1$ -shot defect detection results. GATE-AD localizes more accurately and robustly defects, across multiple pose, placement, and abnormality type settings.

	MVTec AD		VisA		MPDD
Shot–Method	I-AUROC	P-AUROC	I-AUROC	P-AUROC	I-AUROC	P-AUROC
1-shot
GATE-AD	97.7 $\pm$ 0.2	96.3 $\pm$ 0.1	93.5 $\pm$ 0.2	97.7 $\pm$ 0.1	84.3 $\pm$ 1.1	97.4 $\pm$ 0.2
GATE-AD (GCN)	97.3 $\pm$ 0.2	95.9 $\pm$ 0.2	92.7 $\pm$ 0.4	97.4 $\pm$ 0.1	82.1 $\pm$ 0.9	95.7 $\pm$ 0.1
TE	97.1 $\pm$ 0.2	95.7 $\pm$ 0.1	93.4 $\pm$ 0.2	97.5 $\pm$ 0.1	82.4 $\pm$ 0.9	96.3 $\pm$ 0.1
GATE-AD w/o RA	94.1 $\pm$ 0.5	95.1 $\pm$ 0.1	80.5 $\pm$ 0.3	96.0 $\pm$ 0.2	55.7 $\pm$ 1.1	94.4 $\pm$ 0.4

Table 2: FS-IVAD ablation study on the MVTec AD, VisA, and MPDD benchmarks.

Panel of Figure 3 — Figure 3: FS-IVAD detection performance (image-level AUROC) versus per-image inference time (ms), under the $1$ -shot setting on the MVTec AD benchmark.

Appendix figures & tables18 assets

Supplementary material from the paper’s appendix.

Appendix

Panel of Figure 4 — Figure 4: Effect of the input image resolution on the MVTec AD and VisA benchmarks.

Panel of Figure 5 — Figure 5: Effect of the backbone on the MVTec AD, VisA, and MPDD benchmarks.

Panel of Figure 6 — Figure 6: Effect of the number of last transformer layers on the MVTec and VisA benchmarks.

Panel of Figure 7 — Figure 7: Effect of the number of attentional layers on the MVTec AD and VisA benchmarks.

Panel of Figure 8 — Figure 8: Effect of the cardinality of the GAT layers on the MVTec AD and VisA benchmarks.

Panel of Figure 9 — Figure 9: Effect of the cardinality of the latent representation on the MVTec AD and VisA benchmarks.

Panel of Figure 10 — Figure 10: Effect of the objective function on the MVTec AD and VisA benchmarks.

Panel of Figure 11 — Figure 11: Effect of the masking ratio on the MVTec and VisA benchmarks.

Panel of Figure 12 — Figure 12: Effect of the dropout rate ablation on the MVTec and VisA benchmarks.

Panel of Figure 13 — Figure 13: Effect of the anomalous patches ratio on the MVTec AD and VisA benchmarks.

Object	1-shot		2-shot		4-shot		8-shot
Object	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO
Image AUROC (%)
Bottle	100.0 $\pm$ 0.0	99.7 $\pm$ 0.2	99.8 $\pm$ 0.1	99.9 $\pm$ 0.1	100.0 $\pm$ 0.0	99.9 $\pm$ 0.1	100.0 $\pm$ 0.0	100.0 $\pm$ 0.0
Cable	94.7 $\pm$ 0.1	92.7 $\pm$ 0.8	95.5 $\pm$ 0.1	92.4 $\pm$ 1.1	94.3 $\pm$ 0.1	93.8 $\pm$ 0.9	95.0 $\pm$ 0.4	95.2 $\pm$ 0.3
Capsule	93.1 $\pm$ 0.1	90.2 $\pm$ 5.5	94.3 $\pm$ 0.7	89.2 $\pm$ 7.9	95.0 $\pm$ 0.9	95.8 $\pm$ 0.5	95.8 $\pm$ 0.8	95.6 $\pm$ 0.5
Carpet	99.8 $\pm$ 0.0	100.0 $\pm$ 0.0	99.9 $\pm$ 0.0	100.0 $\pm$ 0.0	99.8 $\pm$ 0.0	100.0 $\pm$ 0.0	99.9 $\pm$ 0.0	100.0 $\pm$ 0.0

Appendix

Table 3: FS-IVAD object-level detection results for GATE-AD and AnomalyDINO on the MVTec AD benchmark. Both methods use the standardized sample selection process (Seed:

0

Object	1-shot		2-shot		4-shot		8-shot
Object	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO
Image AUROC (%)
Candle	97.8 $\pm$ 0.4	87.9 $\pm$ 0.3	97.6 $\pm$ 0.3	89.4 $\pm$ 3.0	97.5 $\pm$ 0.3	91.3 $\pm$ 2.9	97.7 $\pm$ 0.2	93.5 $\pm$ 1.2
Capsules	99.3 $\pm$ 0.3	98.4 $\pm$ 0.5	99.5 $\pm$ 0.2	98.9 $\pm$ 0.1	99.6 $\pm$ 0.2	99.2 $\pm$ 0.1	99.6 $\pm$ 0.2	99.2 $\pm$ 0.1
Cashew	95.5 $\pm$ 0.4	86.1 $\pm$ 3.6	96.1 $\pm$ 0.2	89.4 $\pm$ 3.8	98.7 $\pm$ 0.2	94.5 $\pm$ 0.7	98.6 $\pm$ 0.2	95.3 $\pm$ 0.6
Chewing gum	98.8 $\pm$ 0.2	98.0 $\pm$ 0.4	99.1 $\pm$ 0.1	98.6 $\pm$ 0.4	99.3 $\pm$ 0.1	98.8 $\pm$ 0.2	99.4 $\pm$ 0.1	98.8 $\pm$ 0.2

Appendix

Table 4: FS-IVAD object-level detection results for GATE-AD and AnomalyDINO on the VisA benchmark. Both methods use the standardized sample selection process (Seed:

0

Object	1-shot		2-shot		4-shot		8-shot
Object	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO	GATE-AD	AnomalyDINO
Image AUROC (%)
Bracket black	83.7 $\pm$ 3.0	51.0 $\pm$ 6.5	83.4 $\pm$ 2.0	59.4 $\pm$ 8.0	80.9 $\pm$ 2.0	64.2 $\pm$ 8.1	85.7 $\pm$ 0.6	69.5 $\pm$ 2.7
Bracket brown	49.7 $\pm$ 1.9	52.0 $\pm$ 4.6	51.7 $\pm$ 1.3	56.8 $\pm$ 4.4	56.1 $\pm$ 2.1	58.6 $\pm$ 1.2	57.0 $\pm$ 2.0	61.3 $\pm$ 2.8
Bracket white	80.8 $\pm$ 3.3	70.8 $\pm$ 2.3	81.1 $\pm$ 2.2	72.0 $\pm$ 2.8	81.0 $\pm$ 1.9	65.7 $\pm$ 16.7	93.3 $\pm$ 1.7	72.5 $\pm$ 20.1
Connector	93.8 $\pm$ 1.0	80.6 $\pm$ 4.1	91.8 $\pm$ 0.8	83.1 $\pm$ 3.8	89.4 $\pm$ 1.0	83.7 $\pm$ 1.7	93.6 $\pm$ 1.3	87.9 $\pm$ 3.1

Appendix

Table 5: FS-IVAD object-level detection results for GATE-AD and AnomalyDINO on the MPDD benchmark. Both methods use the standardized sample selection process (Seed:

0

Figure 14: Indicative $1$ -shot defect detection results on the MVTec AD benchmark.

Figure 15: Indicative $1$ -shot defect detection results on the VisA benchmark.

Figure 16: Indicative $1$ -shot defect detection results on the MPDD benchmark.

Figure 17: Indicative examples of defect detection failure cases across the MVTec AD, VisA, and MPDD benchmarks.

	MVTec AD		VisA		MPDD
Method	Training	Inference	Training	Inference	Training	Inference
GATE-AD	6.7 $\pm$ 0.9	0.03 $\pm$ 0.00	7.8 $\pm$ 0.9	0.04 $\pm$ 0.00	4.8 $\pm$ 0.6	0.03 $\pm$ 0.00
GATE-AD (GCN)	9.0 $\pm$ 3.5	0.03 $\pm$ 0.00	10.4 $\pm$ 3.4	0.04 $\pm$ 0.00	4.4 $\pm$ 0.3	0.03 $\pm$ 0.00
TE	62.9 $\pm$ 16.1	0.04 $\pm$ 0.00	83.1 $\pm$ 23.4	0.06 $\pm$ 0.00	46.6 $\pm$ 10.0	0.05 $\pm$ 0.00
GATE-AD w/o RA	25.4 $\pm$ 4.0	0.04 $\pm$ 0.00	28.4 $\pm$ 3.9	0.05 $\pm$ 0.00	14.8 $\pm$ 0.3	0.04 $\pm$ 0.00
GATE-AD AE	16.1 $\pm$ 6.5	0.04 $\pm$ 0.00	19.0 $\pm$ 6.7	0.05 $\pm$ 0.00	18.6 $\pm$ 1.3	0.04 $\pm$ 0.00

Appendix

Table 6: Per-image training and inference time (ms) on the MVTec AD, VisA, and MPDD benchmarks.

Explore similar work

ConceptADapt: Concept-guided Adaptive Feature Reconstruction with Dynamic Attention for Few-Shot Industrial Anomaly Detection
Aug 6, 2026Yufei Li, Yicheng Ruan, Long Tian +2Few-Shot Anomaly Detection Industrial Anomaly Detection
School of Computer Science and Technology, Xidian University, Xi’an, China
DriftAD: Visually-Guided Text Drift for Few-Shot Industrial Anomaly Detection
Aug 24, 2026Wenyang Liu, Tianyi Liu, Dongshuo Zhang +2Few-Shot Anomaly Detection Industrial Anomaly Detection
Nanyang Technological University Singapore · Huazhong University of Science and Technology, Wuhan, China
SubspaceAD: Training-Free Few-Shot Anomaly Detection via Subspace Modeling
Feb 26, 2026Camile Lendering, Erkut Akdag, Egor BondarevFew-Shot Anomaly Detection Anomaly Detection
AIMS Group, Department of Electrical Engineering, Eindhoven University of Technology

Aug 6, 2026J/K move · Enter open · S save

ConceptADapt: Concept-guided Adaptive Feature Reconstruction with Dynamic Attention for Few-Shot Industrial Anomaly Detection

Yufei Li, Yicheng Ruan, Long Tian, Dongsheng Wang+1

School of Computer Science and Technology, Xidian University, Xi’an, China

Few-Shot Anomaly Detection Industrial Anomaly Detection Anomaly Localization Reconstruction-Based Anomaly Detection

Open paper AskPDF arXiv

Few-shot industrial anomaly detection (FS-IAD) focuses on detecting and localizing visual defects in industrial inspection during the cold-start phase, where only a limited number of normal training samples are available per category. Recent advances in this field predominantly leverage visual features from foundation-model and have achieved promising performance. Despite the strong representational power of foundation-model features, the model generalization remains fragile due to the extreme scarcity of normal training data.To address this pivotal issue, we propose ConceptADapt, a concept-guided adaptive feature reconstruction model with dynamic attention. Specifically, our model pre-learns a set of fixed normal concepts from the limited support features and leverages them to mine relationships with query features, thereby recalibrating their statistics for improved anomaly detection at test time. To mitigate the prevalent feature shortcut problem, which is particularly severe under low-data regimes, we further develop a dynamic attention mechanism integrated with sparse autoencoders to learn robust normal concepts during training. Moreover, to enable fast adaptation during inference, our model remains lightweight by incorporating LoRA into the attention module, which introduces only minimal updating parameters.Extensive experiments on three widely adopted FS-IAD benchmarks, including MVTec-AD, VisA, and MPDD, demonstrate that our model consistently outperforms state-of-the-art (SOTA) approaches across both detection and localization tasks, achieving significant improvements under various shot settings.

Abstract

Explore similar work

ConceptADapt: Concept-guided Adaptive Feature Reconstruction with Dynamic Attention for Few-Shot Industrial Anomaly Detection

DriftAD: Visually-Guided Text Drift for Few-Shot Industrial Anomaly Detection

SubspaceAD: Training-Free Few-Shot Anomaly Detection via Subspace Modeling