Spatially-weighted Anomaly Detection with Regression Model
This work addresses anomaly detection in applications like medical screening and quality checks, offering an incremental improvement by leveraging available anomaly data and enhancing robustness to noise.
The paper tackles the problem of visual anomaly detection by proposing a spatially-weighted reconstruction-loss method that incorporates anomaly hints and regression model likelihood, achieving better performance than other methods on three datasets.
Visual anomaly detection is common in several applications including medical screening and production quality check. Although a definition of the anomaly is an unknown trend in data, in many cases some hints or samples of the anomaly class can be given in advance. Conventional methods cannot use the available anomaly data, and also do not have a robustness of noise. In this paper, we propose a novel spatially-weighted reconstruction-loss-based anomaly detection with a likelihood value from a regression model trained by all known data. The spatial weights are calculated by a region of interest generated from employing visualization of the regression model. We introduce some ways to combine with various strategies to propose a state-of-the-art method. Comparing with other methods on three different datasets, we empirically verify the proposed method performs better than the others.