MECAD: A multi-expert architecture for continual anomaly detection
By: Malihe Dahmardeh, Francesco Setti
Potential Business Impact:
Teaches computers to spot new problems over time.
In this paper we propose MECAD, a novel approach for continual anomaly detection using a multi-expert architecture. Our system dynamically assigns experts to object classes based on feature similarity and employs efficient memory management to preserve the knowledge of previously seen classes. By leveraging an optimized coreset selection and a specialized replay buffer mechanism, we enable incremental learning without requiring full model retraining. Our experimental evaluation on the MVTec AD dataset demonstrates that the optimal 5-expert configuration achieves an average AUROC of 0.8259 across 15 diverse object categories while significantly reducing knowledge degradation compared to single-expert approaches. This framework balances computational efficiency, specialized knowledge retention, and adaptability, making it well-suited for industrial environments with evolving product types.
Similar Papers
CADE: Continual Weakly-supervised Video Anomaly Detection with Ensembles
CV and Pattern Recognition
Helps security cameras learn new dangers without forgetting old ones.
ExpertAD: Enhancing Autonomous Driving Systems with Mixture of Experts
Robotics
Makes self-driving cars safer and faster.
On-Device Continual Learning for Unsupervised Visual Anomaly Detection in Dynamic Manufacturing
Machine Learning (CS)
Finds factory flaws faster on old machines.