mlsanityData to 14 Oct 2026

CoGeoAD: Hierarchical Color-Geometric Fusion with Multi-View Attention for Zero-Shot 3D Anomaly Detection

arXiv 2606.25273 ↗First posted 24 Jun 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Zero-shot 3D anomaly detection is essential for industrial quality inspection, where labeled anomaly samples are scarce. Meanwhile, existing methods lack an effective mechanism to fuse complementary 2D color images with 3D geometric structures, limiting their ability to detect both surface and structural defects in a unified framework. To address these issues, we propose CoGeoAD, a unified CLIP-based framework that fuses color and geometric features by constructing pixel-aligned paired multi-view images. The framework introduces a Data-Driven Multi-View Attention (MVA) mechanism to adaptively aggregate 3D features and a Multi-Stage Color-Geometric Fusion (MS-CGF) module to hierarchically integrate multi-level features from both modalities. Extensive experiments on the MVTec3D-AD and Eyecandies benchmarks demonstrate that CoGeoAD achieves state-of-the-art performance, effectively capturing both structural and textural anomalies in complex industrial scenarios. our source code is available at https://github.com/kingdomShu/CoGeoAD.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 39 of its 44 references that are in the map. First posted on arXiv 24 Jun 2026.

Builds on

39 of 44 references.

References · 39

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2021PointCLIP: Point Cloud Understanding by CLIPcited by 463
  3. 2019Uninformed Students: Student-Teacher Anomaly Detection with Discriminative Latent Embeddingscited by 386
  4. 2023WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentationcited by 315
  5. 2022A Unified Model for Multi-class Anomaly Detectioncited by 268
  6. 2022PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learningcited by 187
  7. 2021The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localizationcited by 178
  8. 2023AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Modelscited by 171
  9. 2023Multimodal Industrial Anomaly Detection via Hybrid Fusioncited by 139
  10. 2024Real-IAD: A Real-World Multi-View Dataset for Benchmarking Versatile Industrial Anomaly Detectioncited by 139
  11. 2022Registration based Few-Shot Anomaly Detectioncited by 133
  12. 2023APRIL-GAN: A Zero-/Few-Shot Anomaly Classification and Segmentation Method for CVPR 2023 VAND Workshop Challenge Tracks 1&2: 1st Place on Zero-shot AD and 4th Place on Few-shot ADcited by 130
27 more
  1. 2024AdaCLIP: Adapting CLIP with Hybrid Learnable Prompts for Zero-Shot Anomaly Detectioncited by 127
  2. 2023Real3D-AD: A Dataset of Point Cloud Anomaly Detectioncited by 87
  3. 2023Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detectioncited by 75
  4. 2024A Survey on Visual Anomaly Detection: Challenge, Approach, and Prospectcited by 69
  5. 2022Deep Learning for Unsupervised Anomaly Localization in Industrial Images: A Surveycited by 68
  6. 2022Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptorscited by 66
  7. 2022The Eyecandies Dataset for Unsupervised Multimodal Anomaly Detection and Localizationcited by 57
  8. 2024VCP-CLIP: A visual context prompting model for zero-shot anomaly segmentationcited by 50
  9. 2024R3D-AD: Reconstruction via Diffusion for 3D Anomaly Detectioncited by 45
  10. 2023EasyNet: An Easy Network for 3D Industrial Anomaly Detectioncited by 43
  11. 2024Look Inside for More: Internal Spatial Modality Perception for 3D Anomaly Detectioncited by 30
  12. 2024PointAD: Comprehending 3D Anomalies from Points and Pixels for Zero-shot 3D Anomaly Detectioncited by 29
  13. 2025AdaptCLIP: Adapting CLIP for Universal Visual Anomaly Detectioncited by 29
  14. 2024SeaS: Few-shot Industrial Anomaly Image Generation with Separation and Sharing Fine-tuningcited by 27
  15. 2025MC3D-AD: A Unified Geometry-aware Reconstruction Model for Multi-category 3D Anomaly Detectioncited by 23
  16. 2025Learning to Detect Multi-class Anomalies with Just One Normal Image Promptcited by 18
  17. 2025Towards High-Resolution 3D Anomaly Detection: A Scalable Dataset and Real-Time Framework for Subtle Industrial Defectscited by 15
  18. 2025Taming Anomalies with Down-Up Sampling Networks: Group Center Preserving Reconstruction for 3D Anomaly Detectioncited by 11
  19. 2026VisualAD: Language-Free Zero-Shot Anomaly Detection via Vision Transformercited by 11
  20. 2024AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarioscited by 9
  21. 2025Bridging 3D Anomaly Localization and Repair via High-Quality Continuous Geometric Representationcited by 9
  22. 2026GS-CLIP: Zero-shot 3D Anomaly Detection by Geometry-Aware Prompt and Synergistic View Representation Learningcited by 4
  23. 2025Soft Reasoning Paths for Knowledge Graph Completioncited by 3
  24. 2025Enhancing Zero-Shot Anomaly Detection: CLIP-SAM Collaboration with Cascaded Promptscited by 3
  25. 2025MuSc-V2: Zero-Shot Multimodal Industrial Anomaly Classification and Segmentation with Mutual Scoring of Unlabeled Samplescited by 3
  26. 2025StackCLIP: Clustering-Driven Stacked Prompt in Zero-Shot Industrial Anomaly Detectioncited by 2
  27. 2025PointAD+: Learning Hierarchical Representations for Zero-shot 3D Anomaly Detectioncited by 1

Built on by

Not cited yet.