# HarmoniAD: Harmonizing Local Structures and Global Semantics for Anomaly Detection

arXiv: https://arxiv.org/abs/2601.00327 · HTML: https://mlsanity.com/paper/2601.00327/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 18 references that are in the map.
- First posted on arXiv 1 Jan 2026.

## Builds on (most cited first, 12 of 12)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Towards Total Recall in Industrial Anomaly Detection](https://mlsanity.com/paper/2106.08265/) (2021)
- [PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localization](https://mlsanity.com/paper/2011.08785/) (2020)
- [SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentation](https://mlsanity.com/paper/2207.14315/) (2022)
- [WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation](https://mlsanity.com/paper/2303.14814/) (2023)
- [AA-CLIP: Enhancing Zero-shot Anomaly Detection via Anomaly-Aware CLIP](https://mlsanity.com/paper/2503.06661/) (2025)
- [SPADE: Semi-supervised Anomaly Detection under Distribution Mismatch](https://mlsanity.com/paper/2212.00173/) (2022)
- [Unlocking the Potential of Reverse Distillation for Anomaly Detection](https://mlsanity.com/paper/2412.07579/) (2024)
- [KAnoCLIP: Zero-Shot Anomaly Detection through Knowledge-Driven Prompt Learning and Enhanced Cross-Modal Integration](https://mlsanity.com/paper/2501.03786/) (2025)
- [Neural Network Training Strategy to Enhance Anomaly Detection Performance: A Perspective on Reconstruction Loss Amplification](https://mlsanity.com/paper/2308.14595/) (2023)
- [Towards Unbiased Evaluation of Time-series Anomaly Detector](https://mlsanity.com/paper/2409.13053/) (2024)
- [Unsupervised Anomaly Detection with Local-Sensitive VQVAE and Global-Sensitive Transformers](https://mlsanity.com/paper/2303.17505/) (2023)

## Abstract (arXiv)

Anomaly detection is crucial in industrial product quality inspection. Failing to detect tiny defects often leads to serious consequences. Existing methods face a structure-semantics trade-off: structure-oriented models (such as frequency-based filters) are noise-sensitive, while semantics-oriented models (such as CLIP-based encoders) often miss fine details. To address this, we propose HarmoniAD, a frequency-guided dual-branch framework. Features are first extracted by the CLIP image encoder, then transformed into the frequency domain, and finally decoupled into high- and low-frequency paths for complementary modeling of structure and semantics. The high-frequency branch is equipped with a fine-grained structural attention module (FSAM) to enhance textures and edges for detecting small anomalies, while the low-frequency branch uses a global structural context module (GSCM) to capture long-range dependencies and preserve semantic consistency. Together, these branches balance fine detail and global semantics. HarmoniAD further adopts a multi-class joint training strategy, and experiments on MVTec-AD, VisA, and BTAD show state-of-the-art performance with both sensitivity and robustness.