Real time dense anomaly detection by learning on synthetic negative data
This work addresses anomaly detection for computer vision applications, but it is incremental as it builds on existing hybrid methods.
The paper tackles dense anomaly detection by extending a hybrid method with a generative flow to sample synthetic negatives at the inlier distribution border, enabling training without real negative data. Experiments analyze the impact of synthetic data and validate the energy-based density's contribution.
Most approaches to dense anomaly detection rely on generative modeling or on discriminative methods that train with negative data. We consider a recent hybrid method that optimizes the same shared representation according to cross-entropy of the discriminative predictions, and negative log likelihood of the predicted energy-based density. We extend that work with a jointly trained generative flow that samples synthetic negatives at the border of the inlier distribution. The proposed extension provides potential to learn the hybrid method without real negative data. Our experiments analyze the impact of training with synthetic negative data and validate contribution of the energy-based density during training and evaluation.