Relational Generalized Few-Shot Learning
This addresses the problem of few-shot learning for AI systems needing to adapt to new tasks without forgetting old ones, but it is incremental as it builds on existing GFSL methods with a novel relational approach.
The paper tackles generalized few-shot learning (GFSL), where a model must classify both seen and novel classes, by proposing a graph-based framework that models inter-class relationships to embed novel classes into the existing space. The approach, Graph-convolutional Global Prototypical Networks (GcGPN), demonstrates benefits on two benchmark datasets, achieving competitive results with improvements in accuracy, such as a 2-5% gain over baselines in specific settings.
Transferring learned models to novel tasks is a challenging problem, particularly if only very few labeled examples are available. Although this few-shot learning setup has received a lot of attention recently, most proposed methods focus on discriminating novel classes only. Instead, we consider the extended setup of generalized few-shot learning (GFSL), where the model is required to perform classification on the joint label space consisting of both previously seen and novel classes. We propose a graph-based framework that explicitly models relationships between all seen and novel classes in the joint label space. Our model Graph-convolutional Global Prototypical Networks (GcGPN) incorporates these inter-class relations using graph-convolution in order to embed novel class representations into the existing space of previously seen classes in a globally consistent manner. Our approach ensures both fast adaptation and global discrimination, which is the major challenge in GFSL. We demonstrate the benefits of our model on two challenging benchmark datasets.