Learning to Adapt to Unseen Abnormal Activities under Weak Supervision

TL;DR

This paper introduces a meta-learning framework for weakly supervised video anomaly detection, enabling better generalization to unseen abnormal activities with only video-level labels, and demonstrates improved localization performance on challenging datasets.

Contribution

The work proposes a novel meta-learning approach that enhances anomaly detectors' ability to adapt to unseen events under weak supervision, addressing generalization issues of prior methods.

Findings

Improved localization of unseen abnormal events on UCF-Crime and ShanghaiTech datasets.

Meta-learning scheme leads to better initialization for anomaly detection models.

Annotated missing labels in UCF-Crime for more effective evaluation.

Abstract

We present a meta-learning framework for weakly supervised anomaly detection in videos, where the detector learns to adapt to unseen types of abnormal activities effectively when only video-level annotations of binary labels are available. Our work is motivated by the fact that existing methods suffer from poor generalization to diverse unseen examples. We claim that an anomaly detector equipped with a meta-learning scheme alleviates the limitation by leading the model to an initialization point for better optimization. We evaluate the performance of our framework on two challenging datasets, UCF-Crime and ShanghaiTech. The experimental results demonstrate that our algorithm boosts the capability to localize unseen abnormal events in a weakly supervised setting. Besides the technical contributions, we perform the annotation of missing labels in the UCF-Crime dataset and make our task evaluated effectively.