Taxonomic Class Incremental Learning
It addresses a more realistic continual learning setup for AI systems by mimicking human taxonomic curricula, though it is incremental as it builds on existing approaches.
The paper tackles the problem of continual learning by proposing Taxonomic Class Incremental Learning (TCIL), which organizes tasks based on a taxonomic class tree rather than random classes, and shows it outperforms state-of-the-art methods with 2% higher final accuracy on CIFAR-100 and 3% on ImageNet-100.
The problem of continual learning has attracted rising attention in recent years. However, few works have questioned the commonly used learning setup, based on a task curriculum of random class. This differs significantly from human continual learning, which is guided by taxonomic curricula. In this work, we propose the Taxonomic Class Incremental Learning (TCIL) problem. In TCIL, the task sequence is organized based on a taxonomic class tree. We unify existing approaches to CIL and taxonomic learning as parameter inheritance schemes and introduce a new such scheme for the TCIL learning. This enables the incremental transfer of knowledge from ancestor to descendant class of a class taxonomy through parameter inheritance. Experiments on CIFAR-100 and ImageNet-100 show the effectiveness of the proposed TCIL method, which outperforms existing SOTA methods by 2% in terms of final accuracy on CIFAR-100 and 3% on ImageNet-100.