In Defense of the Learning Without Forgetting for Task Incremental Learning
This work addresses catastrophic forgetting in continual learning systems, which is a critical problem for AI systems that need to learn from online streams of tasks, though it is incremental as it improves an existing method rather than introducing a new paradigm.
The paper challenges the prevailing view that Learning Without Forgetting (LwF) fails to scale to long task sequences in continual learning, showing that with the right architecture and standard augmentations, LwF surpasses the latest algorithms in task incremental scenarios, as demonstrated by extensive experiments on CIFAR-100 and Tiny-ImageNet.
Catastrophic forgetting is one of the major challenges on the road for continual learning systems, which are presented with an on-line stream of tasks. The field has attracted considerable interest and a diverse set of methods have been presented for overcoming this challenge. Learning without Forgetting (LwF) is one of the earliest and most frequently cited methods. It has the advantages of not requiring the storage of samples from the previous tasks, of implementation simplicity, and of being well-grounded by relying on knowledge distillation. However, the prevailing view is that while it shows a relatively small amount of forgetting when only two tasks are introduced, it fails to scale to long sequences of tasks. This paper challenges this view, by showing that using the right architecture along with a standard set of augmentations, the results obtained by LwF surpass the latest algorithms for task incremental scenario. This improved performance is demonstrated by an extensive set of experiments over CIFAR-100 and Tiny-ImageNet, where it is also shown that other methods cannot benefit as much from similar improvements.