MetaMix: Improved Meta-Learning with Interpolation-based Consistency Regularization
This work addresses the challenge of forming generalizable decision boundaries in few-shot learning for researchers and practitioners, representing an incremental improvement over existing MAML-based methods.
The paper tackles the problem of limited generalizability in decision boundaries for Model-Agnostic Meta-Learning (MAML) methods in few-shot classification by proposing MetaMix, an interpolation-based consistency regularization approach that generates virtual feature-target pairs within episodes. Experiments on mini-ImageNet, CUB, and FC100 datasets show that MetaMix improves MAML-based algorithms and achieves state-of-the-art results when integrated with Meta-Transfer Learning.
Model-Agnostic Meta-Learning (MAML) and its variants are popular few-shot classification methods. They train an initializer across a variety of sampled learning tasks (also known as episodes) such that the initialized model can adapt quickly to new tasks. However, current MAML-based algorithms have limitations in forming generalizable decision boundaries. In this paper, we propose an approach called MetaMix. It generates virtual feature-target pairs within each episode to regularize the backbone models. MetaMix can be integrated with any of the MAML-based algorithms and learn the decision boundaries generalizing better to new tasks. Experiments on the mini-ImageNet, CUB, and FC100 datasets show that MetaMix improves the performance of MAML-based algorithms and achieves state-of-the-art result when integrated with Meta-Transfer Learning.