PTN: A Poisson Transfer Network for Semi-supervised Few-shot Learning
This work addresses the problem of effectively leveraging unlabeled data in semi-supervised few-shot learning, which is critical for researchers and practitioners dealing with limited labeled data in new classes.
This paper introduces the Poisson Transfer Network (PTN) to enhance semi-supervised few-shot learning by effectively utilizing unlabeled data. PTN employs a Poisson Merriman Bence Osher (MBO) model for stable label propagation and uses contrastive learning with unlabeled samples to transfer knowledge from base to novel classes, thereby improving embedding generality. The method achieves state-of-the-art performance on miniImageNet and tieredImageNet datasets.
The predicament in semi-supervised few-shot learning (SSFSL) is to maximize the value of the extra unlabeled data to boost the few-shot learner. In this paper, we propose a Poisson Transfer Network (PTN) to mine the unlabeled information for SSFSL from two aspects. First, the Poisson Merriman Bence Osher (MBO) model builds a bridge for the communications between labeled and unlabeled examples. This model serves as a more stable and informative classifier than traditional graph-based SSFSL methods in the message-passing process of the labels. Second, the extra unlabeled samples are employed to transfer the knowledge from base classes to novel classes through contrastive learning. Specifically, we force the augmented positive pairs close while push the negative ones distant. Our contrastive transfer scheme implicitly learns the novel-class embeddings to alleviate the over-fitting problem on the few labeled data. Thus, we can mitigate the degeneration of embedding generality in novel classes. Extensive experiments indicate that PTN outperforms the state-of-the-art few-shot and SSFSL models on miniImageNet and tieredImageNet benchmark datasets.