ReMP: Rectified Metric Propagation for Few-Shot Learning
This work addresses the problem of improving generalization from limited examples in few-shot learning for researchers and practitioners, representing an incremental advancement in metric-based methods.
This paper identifies that a rectified metric space, which maintains metric consistency from training to testing, is crucial for metric-based few-shot learning. By modifying the objective and optimizing an attentive prototype propagation network with a repulsive force, their method, ReMP, achieves state-of-the-art performance on various few-shot learning datasets.
Few-shot learning features the capability of generalizing from a few examples. In this paper, we first identify that a discriminative feature space, namely a rectified metric space, that is learned to maintain the metric consistency from training to testing, is an essential component to the success of metric-based few-shot learning. Numerous analyses indicate that a simple modification of the objective can yield substantial performance gains. The resulting approach, called rectified metric propagation (ReMP), further optimizes an attentive prototype propagation network, and applies a repulsive force to make confident predictions. Extensive experiments demonstrate that the proposed ReMP is effective and efficient, and outperforms the state of the arts on various standard few-shot learning datasets.