InDiD: Instant Disorder Detection via Representation Learning
This work addresses the need for fast disorder detection in applications like video surveillance and sensor monitoring, offering an incremental improvement over existing methods.
The paper tackles the problem of detecting change points in sequential data, such as videos and sensor streams, by proposing a loss function that balances detection delay and false alarms, enabling representation learning for deep models. It achieves an F1 score of 0.53 for explosion detection in video, outperforming baselines with scores of 0.31 and 0.35.
For sequential data, a change point is a moment of abrupt regime switch in data streams. Such changes appear in different scenarios, including simpler data from sensors and more challenging video surveillance data. We need to detect disorders as fast as possible. Classic approaches for change point detection (CPD) might underperform for semi-structured sequential data because they cannot process its structure without a proper representation. We propose a principled loss function that balances change detection delay and time to a false alarm. It approximates classic rigorous solutions but is differentiable and allows representation learning for deep models. We consider synthetic sequences, real-world data sensors and videos with change points. We carefully labelled available data with change point moments for video data and released it for the first time. Experiments suggest that complex data require meaningful representations tailored for the specificity of the CPD task -- and our approach provides them outperforming considered baselines. For example, for explosion detection in video, the F1 score for our method is 0.53 compared to baseline scores of 0.31 and 0.35.