Cleaning Label Noise with Clusters for Minimally Supervised Anomaly Detection
This work addresses the challenge of noisy video-level labels for anomaly detection, which is crucial for applications like surveillance, but it is incremental as it builds on existing weakly supervised methods.
The paper tackles the problem of label noise in weakly supervised anomaly detection in videos by proposing a method that uses binary clustering to clean noisy labels, achieving 78.27% and 84.16% frame-level AUC on UCF-crime and ShanghaiTech datasets.
Learning to detect real-world anomalous events using video-level annotations is a difficult task mainly because of the noise present in labels. An anomalous labelled video may actually contain anomaly only in a short duration while the rest of the video can be normal. In the current work, we formulate a weakly supervised anomaly detection method that is trained using only video-level labels. To this end, we propose to utilize binary clustering which helps in mitigating the noise present in the labels of anomalous videos. Our formulation encourages both the main network and the clustering to complement each other in achieving the goal of weakly supervised training. The proposed method yields 78.27% and 84.16% frame-level AUC on UCF-crime and ShanghaiTech datasets respectively, demonstrating its superiority over existing state-of-the-art algorithms.