mlsanityData to 13 Oct 2026

AT3D-AD: Anomaly Type-Aware 3D Anomaly Detection via Hierarchical Point-Language Alignment

arXiv 2609.25930 ↗First posted 22 Sep 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Detecting and localizing 3D point-cloud defects is essential for industrial inspection. However, existing methods often suffer from imprecise localization due to the lack of anomaly supervision and reliance on single-granularity representations. To address these limitations, we propose Anomaly Type-Aware 3D Anomaly Detection (AT3D-AD), a unified framework for joint detection, localization, and classification. Specifically, we first design the Physics-Driven Parametric Anomaly Synthesis (PDPAS) module employing multiple parametric functions to generate synthetic anomalies, providing explicit anomaly supervision. Then, we propose the Hierarchical Global-Local Anomaly Alignment (HiGLA) module to align global and local representations within the normal and anomalous groups. Finally, we propose the Semantic-Geometric Anomaly Classification (SGAC) module to jointly learn localization and classification, yielding spatially precise and type-discriminative anomaly representations. Extensive experiments establish new state-of-the-art performance on all four benchmarks. AT3D-AD achieves Object/Point AUROC scores of 98.1%/98.9% on Anomaly-ShapeNet and 95.0%/95.2% on Real3D-AD, while reaching 74.2% Macro-F1 for anomaly-type recognition on Real3D-AD.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Builds on 21 of its 28 references that are in the map. First posted on arXiv 22 Sep 2026.

Builds on

21 of 28 references.

References · 21

  1. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,389
  2. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,222
  3. 2021Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modelingcited by 648
  4. 2021PointCLIP: Point Cloud Understanding by CLIPcited by 463
  5. 2022ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understandingcited by 265
  6. 2023ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understandingcited by 199
  7. 2022PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learningcited by 187
  8. 2023Multimodal Industrial Anomaly Detection via Hybrid Fusioncited by 138
  9. 2022Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detectioncited by 92
  10. 2023Real3D-AD: A Dataset of Point Cloud Anomaly Detectioncited by 87
  11. 2023Towards Scalable 3D Anomaly Detection and Localization: A Benchmark via 3D Anomaly Synthesis and A Self-Supervised Learning Networkcited by 67
  12. 2024R3D-AD: Reconstruction via Diffusion for 3D Anomaly Detectioncited by 45
9 more
  1. 2024PointAD: Comprehending 3D Anomalies from Points and Pixels for Zero-shot 3D Anomaly Detectioncited by 29
  2. 2024CLIP3D-AD: Extending CLIP for 3D Few-Shot Anomaly Detection with Multi-View Images Generationcited by 16
  3. 2025Boosting Global-Local Feature Matching via Anomaly Synthesis for Multi-Class Point Cloud Anomaly Detectioncited by 16
  4. 2025Towards High-Resolution 3D Anomaly Detection: A Scalable Dataset and Real-Time Framework for Subtle Industrial Defectscited by 15
  5. 2025CASL: Curvature-Augmented Self-supervised Learning for 3D Anomaly Detectioncited by 5
  6. 2025Exploiting Point-Language Models with Dual-Prompts for 3D Anomaly Detectioncited by 4
  7. 2026A Semantically Disentangled Unified Model for Multi-category 3D Anomaly Detectioncited by 3
  8. 2026Back to Point: Exploring Point-Language Models for Zero-Shot 3D Anomaly Detectioncited by 2
  9. 2026Physics-inspired Pseudo Anomaly Generation and Prototype Feature Guidance for 3D Anomaly Detectioncited by 1

Built on by

Not cited yet.