Non-Generative Energy Based Models
This work addresses efficiency and scalability problems for researchers and practitioners using EBMs in computer vision, offering a more practical alternative with significant performance gains.
The paper tackles the computational complexity and stability issues of traditional energy-based models (EBMs) by proposing a non-generative training approach called NG-EBM, which retains benefits like improved calibration, out-of-distribution detection, and adversarial resistance while reducing overhead, achieving a 2.5x improvement in Expected Calibration Error for CIFAR10 and 7.5x for CIFAR100 compared to traditional models.
Energy-based models (EBM) have become increasingly popular within computer vision. EBMs bring a probabilistic approach to training deep neural networks (DNN) and have been shown to enhance performance in areas such as calibration, out-of-distribution detection, and adversarial resistance. However, these advantages come at the cost of estimating input data probabilities, usually using a Langevin based method such as Stochastic Gradient Langevin Dynamics (SGLD), which bring additional computational costs, require parameterization, caching methods for efficiency, and can run into stability and scaling issues. EBMs use dynamical methods to draw samples from the probability density function (PDF) defined by the current state of the network and compare them to the training data using a maximum log likelihood approach to learn the correct PDF. We propose a non-generative training approach, Non-Generative EBM (NG-EBM), that utilizes the {\it{Approximate Mass}}, identified by Grathwohl et al., as a loss term to direct the training. We show that our NG-EBM training strategy retains many of the benefits of EBM in calibration, out-of-distribution detection, and adversarial resistance, but without the computational complexity and overhead of the traditional approaches. In particular, the NG-EBM approach improves the Expected Calibration Error by a factor of 2.5 for CIFAR10 and 7.5 times for CIFAR100, when compared to traditionally trained models.