Curriculum Meta-Learning for Few-shot Classification
This work addresses the problem of improving few-shot learning efficiency for AI systems, but it is incremental as it builds on existing meta-learning methods.
The paper tackles few-shot classification by adapting curriculum training to meta-learning, starting with larger support sets and progressively reducing them, which boosts learning efficiency and generalization, achieving significant gains on two image classification tasks with MAML.
We propose an adaptation of the curriculum training framework, applicable to state-of-the-art meta learning techniques for few-shot classification. Curriculum-based training popularly attempts to mimic human learning by progressively increasing the training complexity to enable incremental concept learning. As the meta-learner's goal is learning how to learn from as few samples as possible, the exact number of those samples (i.e. the size of the support set) arises as a natural proxy of a given task's difficulty. We define a simple yet novel curriculum schedule that begins with a larger support size and progressively reduces it throughout training to eventually match the desired shot-size of the test setup. This proposed method boosts the learning efficiency as well as the generalization capability. Our experiments with the MAML algorithm on two few-shot image classification tasks show significant gains with the curriculum training framework. Ablation studies corroborate the independence of our proposed method from the model architecture as well as the meta-learning hyperparameters