Tal Reiss, George Kour, Naama Zwerdling, Ateret Anaby-Tavor, Yedid Hoshen
When first deploying an anomaly detection system, e.g., to detect out-of-scope queries in chatbots, there are no observed data, making data-driven approaches ineffective. Zero-shot anomaly detection methods offer a solution to such "cold-start" cases, but unfortunately they are often not accurate enough. This paper studies the realistic but underexplored cold-start setting where an anomaly detection model is initialized using zero-shot guidance, but subsequently receives a small number of contaminated observations (namely, that may include anomalies). The goal is to make efficient use of both the zero-shot guidance and the observations. We propose ColdFusion, a method that effectively adapts the zero-shot anomaly detector to contaminated observations. To support future development of this new setting, we propose an evaluation suite consisting of evaluation protocols and metrics.
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Anomaly Detection | BANKING77-OOS | AUC 10% | 81.7 | ColdFusion |
| Anomaly Detection | BANKING77-OOS | AUC 10% | 78.9 | ZS |
| Anomaly Detection | BANKING77-OOS | AUC 10% | 76.7 | DN2 |