Efficient Automatic Meta Optimization Search for Few-Shot Learning

Abstract

Previous works on meta-learning either relied on elaborately hand-designed network structures or adopted specialized learning rules to a particular domain. We propose a universal framework to optimize the meta-learning process automatically by adopting neural architecture search technique (NAS). NAS automatically generates and evaluates meta-learner's architecture for few-shot learning problems, while the meta-learner uses meta-learning algorithm to optimize its parameters based on the distribution of learning tasks. Parameter sharing and experience replay are adopted to accelerate the architectures searching process, so it takes only 1-2 GPU days to find good architectures. Extensive experiments on Mini-ImageNet and Omniglot show that our algorithm excels in few-shot learning tasks. The best architecture found on Mini-ImageNet achieves competitive results when transferred to Omniglot, which shows the high transferability of architectures among different computer vision problems.

References in corpus (9)

• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Practical Bayesian Optimization of Machine Learning Algorithms
• Neural Architecture Search with Reinforcement Learning
• Prioritized Experience Replay
• Efficient Neural Architecture Search via Parameter Sharing
• One-shot Learning with Memory-Augmented Neural Networks
• CMA-ES for Hyperparameter Optimization of Deep Neural Networks
• Hyperparameter Optimization: A Spectral Approach
• Object-Level Representation Learning for Few-Shot Image Classification