How to train your MAML

The field of few-shot learning has recently seen substantial advancements. Most of these advancements came from casting few-shot learning as a meta-learning problem. Model Agnostic Meta Learning or MAML is currently one of the best approaches for few-shot learning via meta-learning. MAML is simple, elegant and very powerful, however, it has a variety of issues, such as being very sensitive to neural network architectures, often leading to instability during training, requiring arduous hyperparameter searches to stabilize training and achieve high generalization and being very computationally expensive at both training and inference times. In this paper, we propose various modifications to MAML that not only stabilize the system, but also substantially improve the generalization performance, convergence speed and computational overhead of MAML, which we call MAML++.

PDF Abstract ICLR 2019 PDF ICLR 2019 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Few-Shot Image Classification Mini-Imagenet 5-way (1-shot) MAML++ Accuracy 52.40 # 85
Few-Shot Image Classification Mini-Imagenet 5-way (5-shot) MAML++ Accuracy 67.15 # 82
Few-Shot Image Classification OMNIGLOT - 1-Shot, 20-way MAML++ Accuracy 97.65 # 6
Few-Shot Image Classification OMNIGLOT - 1-Shot, 5-way MAML++ Accuracy 99.47 # 7
Few-Shot Image Classification OMNIGLOT - 5-Shot, 20-way MAML++ Accuracy 99.33% # 5
Few-Shot Image Classification OMNIGLOT - 5-Shot, 5-way MAML++ Accuracy 99.85% # 6

Methods