Multimodal Model-Agnostic Meta-Learning via Task-Aware Modulation

TL;DR

This paper introduces MMAML, a multimodal extension of MAML, enabling rapid adaptation to diverse task modes by modulating meta-learned priors, improving performance across various few-shot learning domains.

Contribution

The paper proposes a novel multimodal MAML framework that identifies task modes and modulates priors for better adaptation, addressing limitations of shared initializations.

Findings

Effective modulation of meta-learned priors based on task mode

Improved adaptation in few-shot learning tasks

Training on multimodal distributions enhances performance

Abstract

Model-agnostic meta-learners aim to acquire meta-learned parameters from similar tasks to adapt to novel tasks from the same distribution with few gradient updates. With the flexibility in the choice of models, those frameworks demonstrate appealing performance on a variety of domains such as few-shot image classification and reinforcement learning. However, one important limitation of such frameworks is that they seek a common initialization shared across the entire task distribution, substantially limiting the diversity of the task distributions that they are able to learn from. In this paper, we augment MAML with the capability to identify the mode of tasks sampled from a multimodal task distribution and adapt quickly through gradient updates. Specifically, we propose a multimodal MAML (MMAML) framework, which is able to modulate its meta-learned prior parameters according to the identified mode, allowing more efficient fast adaptation. We evaluate the proposed model on a diverse set of few-shot learning tasks, including regression, image classification, and reinforcement learning. The results not only demonstrate the effectiveness of our model in modulating the meta-learned prior in response to the characteristics of tasks but also show that training on a multimodal distribution can produce an improvement over unimodal training.