Task-agnostic Continual Learning with Hybrid Probabilistic ModelsDownload PDF

Published: 15 Jun 2021, Last Modified: 05 May 2023INNF+ 2021 spotlighttalkReaders: Everyone
Keywords: continual learning, normalizing flows, generative models
TL;DR: A normalizing flow based model which continually learns to generate and classify tasks without known task boundaries.
Abstract: Learning new tasks continuously without forgetting on a constantly changing data distribution is essential for real-world problems but extremely challenging for modern deep learning. In this work we propose HCL, a Hybrid generative-discriminative approach to Continual Learning for classification. We model the distribution of each task and each class with a normalizing flow. The flow is used to learn the data distribution, perform classification, identify task changes and avoid forgetting, all leveraging the invertibility and exact likelihood which are uniquely enabled by the normalizing flow model. We use the generative capabilities of the flow to avoid catastrophic forgetting through generative replay and a novel functional regularization technique. For task identification, we use state-of-the-art anomaly detection techniques based on measuring the typicality of model's statistics. We demonstrate the strong performance of HCL on a range of continual learning benchmarks such as split-MNIST, split-CIFAR and SVHN-MNIST.
4 Replies