Energy-Based Models for Anomaly Detection: A Manifold Diffusion Recovery Approach
This work addresses anomaly detection for applications with complex data, though it appears incremental as it builds on existing EBM training methods.
The authors tackled the problem of training energy-based models for anomaly detection by introducing a method that leverages low-dimensional data structures, resulting in strong performance across diverse data types like images, vectors, and acoustic signals.
We present a new method of training energy-based models (EBMs) for anomaly detection that leverages low-dimensional structures within data. The proposed algorithm, Manifold Projection-Diffusion Recovery (MPDR), first perturbs a data point along a low-dimensional manifold that approximates the training dataset. Then, EBM is trained to maximize the probability of recovering the original data. The training involves the generation of negative samples via MCMC, as in conventional EBM training, but from a different distribution concentrated near the manifold. The resulting near-manifold negative samples are highly informative, reflecting relevant modes of variation in data. An energy function of MPDR effectively learns accurate boundaries of the training data distribution and excels at detecting out-of-distribution samples. Experimental results show that MPDR exhibits strong performance across various anomaly detection tasks involving diverse data types, such as images, vectors, and acoustic signals.