mlsanityData to 14 Oct 2026

2D_3D Feature Fusion via Cross-Modal Latent Synthesis and Attention Guided Restoration for Industrial Anomaly Detection

arXiv 2510.21793 ↗First posted 20 Oct 2025cs.CV cs.AI eess.IVcited by 4

Not on the map. Why?

Abstract (arXiv)

Industrial anomaly detection (IAD) increasingly benefits from integrating 2D and 3D data, but robust cross-modal fusion remains challenging. We propose a novel unsupervised framework, Multi-Modal Attention-Driven Fusion Restoration (MAFR), which synthesises a unified latent space from RGB images and point clouds using a shared fusion encoder, followed by attention-guided, modality-specific decoders. Anomalies are localised by measuring reconstruction errors between input features and their restored counterparts. Evaluations on the MVTec 3D-AD and Eyecandies benchmarks demonstrate that MAFR achieves state-of-the-art results, with a mean I-AUROC of 0.972 and 0.901, respectively. The framework also exhibits strong performance in few-shot learning settings, and ablation studies confirm the critical roles of the fusion architecture and composite loss. MAFR offers a principled approach for fusing visual and geometric information, advancing the robustness and accuracy of industrial anomaly detection. Code is available at https://github.com/adabrh/MAFR

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 27 of its 34 references that are in the map. First posted on arXiv 20 Oct 2025.

Builds on

27 of 34 references.

References · 27

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  4. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,401
  5. 2016Wide Residual Networkscited by 3,070
  6. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  7. 2020Point Transformercited by 1,470
  8. 2021Towards Total Recall in Industrial Anomaly Detectioncited by 776
  9. 2019Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detectioncited by 564
  10. 2022Masked Autoencoders for Point Cloud Self-supervised Learningcited by 459
  11. 2020Multiresolution Knowledge Distillation for Anomaly Detectioncited by 285
  12. 2020Same Same But DifferNet: Semi-Supervised Defect Detection with Normalizing Flowscited by 212
15 more
  1. 2021The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localizationcited by 178
  2. 2023Deep Industrial Image Anomaly Detection: A Surveycited by 148
  3. 2023Multimodal Industrial Anomaly Detection via Hybrid Fusioncited by 139
  4. 2022Asymmetric Student-Teacher Networks for Industrial Anomaly Detectioncited by 117
  5. 2022Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detectioncited by 93
  6. 2022DSR -- A dual subspace re-projection network for surface anomaly detectioncited by 90
  7. 2022The Eyecandies Dataset for Unsupervised Multimodal Anomaly Detection and Localizationcited by 57
  8. 2023Anomaly Detection with Conditioned Denoising Diffusion Modelscited by 56
  9. 2020DFR: Deep Feature Reconstruction for Unsupervised Anomaly Segmentationcited by 54
  10. 2022Surface Reconstruction from Point Clouds by Learning Predictive Context Priorscited by 54
  11. 2023Multimodal Industrial Anomaly Detection by Crossmodal Feature Mappingcited by 42
  12. 2022CAINNFlow: Convolutional block Attention modules and Invertible Neural Networks Flow for anomaly detection and localization taskscited by 8
  13. 2024DAS3D: Dual-modality Anomaly Synthesis for 3D Anomaly Detectioncited by 5
  14. 2024Locally-Focused Face Representation for Sketch-to-Image Generation Using Noise-Induced Refinementcited by 3
  15. 2024An Improved Fault Diagnosis Strategy for Induction Motors Using Weighted Probability Ensemble Deep Learningcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026Test-Time Adaptation for Anomaly Segmentation via Topology-Aware Optimal Transport Chainingcited by 0
  2. 2026CMDS-AD: Cross-Modal Dual-Stream Decoupling for Few-Shot Anomaly Detectioncited by 0
  3. 2026Learning Topology-Aware Representations via Test-Time Adaptation for Anomaly Segmentationcited by 0
  4. 2026Multi-Modal Anomaly Detection: A Surveycited by 0