Multi-Scale Memory Comparison for Zero-/Few-Shot Anomaly Detection
This work addresses anomaly detection for industrial defect detection with minimal data, but it is incremental as it builds on existing memory-based techniques.
The paper tackles the challenge of zero-/few-shot anomaly detection in complex industrial scenarios with multiple objects by proposing a multi-scale memory comparison framework, achieving 4th place in the zero-shot track and 2nd place in the few-shot track of the VAND competition.
Anomaly detection has gained considerable attention due to its broad range of applications, particularly in industrial defect detection. To address the challenges of data collection, researchers have introduced zero-/few-shot anomaly detection techniques that require minimal normal images for each category. However, complex industrial scenarios often involve multiple objects, presenting a significant challenge. In light of this, we propose a straightforward yet powerful multi-scale memory comparison framework for zero-/few-shot anomaly detection. Our approach employs a global memory bank to capture features across the entire image, while an individual memory bank focuses on simplified scenes containing a single object. The efficacy of our method is validated by its remarkable achievement of 4th place in the zero-shot track and 2nd place in the few-shot track of the Visual Anomaly and Novelty Detection (VAND) competition.