Abstract:Few-shot learning enables rapid generalization from extremely limited training examples. While previous efforts have utilized meta-learning or data augmentation methods to mitigate the problem of data scarcity, such approaches may struggle to maintain robustness and generalize effectively due to overfitting and noise sensitivity. In this article, we propose a novel approach, the semi-supervised label correction method for few-shot learning (SSLC-FSL), which leverages the data distribution of readily available and easily obtainable unlabeled data. SSLC-FSL iteratively corrects the labels of testing samples with alternating steps of pseudo-labeling and sample selection. The objective of pseudo-labeling is to repurpose graph-based semi-supervised learning for joint prediction of the entire testing set. We then introduce a modulation selection network (MSN) to rank testing samples by learning with noisy labels. The training set is expanded by selecting confident pseudo-labeled samples. In the MSN, a modulation aggregation layer is designed to encode support class information into each testing sample, thereby highlighting target category features and mitigating the negative impact of incorrect labels. The iterative label correction process is repeated until all testing samples are recalled to the expanded support set. To boost the SSLC-FSL algorithm, we pretrain a feature extractor to produce general-purpose representations. Particularly, we investigate two types of auxiliary tasks and their collaborative learning to acquire transferable visual information via an end-to-end multitask learning model. Our SSLC-FSL outperforms current state-of-the-art methods in any shot and all data settings, with up to +27.74% on standard remote sensing benchmarks and +5.70% on standard natural scene benchmarks.

FLAP: A Semi-Supervised Few-Shot Learning Algorithm via Pseudo-Labeling

Domain-Adaptive Few-Shot Learning

Semi-Supervised Few-Shot Classification With Multitask Learning and Iterative Label Correction

A Few-shot-learning-based Method to Object Recognition in Multiple Scenarios.

Few-Shot Learning As Domain Adaptation: Algorithm and Analysis

Adaptive Parametric Prototype Learning for Cross-Domain Few-Shot Classification

Self-Adaptive Label Augmentation for Semi-supervised Few-shot Classification

An Embarrassingly Simple Approach to Semi-Supervised Few-Shot Learning

Co-Learning for Few-Shot Learning

Federated Few-Shot Learning for Mobile NLP.

Pseudo-loss Confidence Metric for Semi-supervised Few-shot Learning

Few-Shot Partial-Label Learning.

Meta-Prototypical Learning for Domain-Agnostic Few-Shot Recognition

Few-shot Learning with Long-Tailed Labels

Towards Cross-Granularity Few-Shot Learning: Coarse-to-Fine Pseudo-Labeling with Visual-Semantic Meta-Embedding

Pseudo-Labeling Based Practical Semi-Supervised Meta-Training for Few-Shot Learning

Few-shot Class-Incremental Semantic Segmentation via Pseudo-Labeling and Knowledge Distillation

ENHANCING PROTOTYPICAL FEW-SHOT LEARNING BY LEVERAGING THE LOCAL-LEVEL STRATEGY

Improved prototypical network for active few-shot learning

Federated Few-Shot Learning with Adversarial Learning