HarmoniAD: Harmonizing Local Structures and Global Semantics for Anomaly Detection
This addresses the problem of detecting tiny defects in industrial product quality inspection, which is incremental as it builds on existing methods like CLIP-based encoders and frequency-based filters.
The paper tackled the structure-semantics trade-off in anomaly detection for industrial inspection by proposing HarmoniAD, a frequency-guided dual-branch framework that balances fine details and global semantics, achieving state-of-the-art performance on datasets like MVTec-AD, VisA, and BTAD.
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.