Wandering Within a World: Online Contextualized Few-Shot Learning
This work addresses the gap between human and machine learning environments for researchers in continual and few-shot learning, though it is incremental as it builds upon existing few-shot methods.
The paper tackles the problem of adapting few-shot learning to an online, continual setting where models learn and evaluate novel classes simultaneously, proposing a new dataset based on indoor imagery and a contextual prototypical memory model that leverages spatiotemporal context, resulting in improved performance through context inference.
We aim to bridge the gap between typical human and machine-learning environments by extending the standard framework of few-shot learning to an online, continual setting. In this setting, episodes do not have separate training and testing phases, and instead models are evaluated online while learning novel classes. As in the real world, where the presence of spatiotemporal context helps us retrieve learned skills in the past, our online few-shot learning setting also features an underlying context that changes throughout time. Object classes are correlated within a context and inferring the correct context can lead to better performance. Building upon this setting, we propose a new few-shot learning dataset based on large scale indoor imagery that mimics the visual experience of an agent wandering within a world. Furthermore, we convert popular few-shot learning approaches into online versions and we also propose a new contextual prototypical memory model that can make use of spatiotemporal contextual information from the recent past.