Unsupervised Meta-Learning For Few-Shot Image Classification

Abstract

Few-shot or one-shot learning of classifiers requires a significant inductive bias towards the type of task to be learned. One way to acquire this is by meta-learning on tasks similar to the target task. In this paper, we propose UMTRA, an algorithm that performs unsupervised, model-agnostic meta-learning for classification tasks. The meta-learning step of UMTRA is performed on a flat collection of unlabeled images. While we assume that these images can be grouped into a diverse set of classes and are relevant to the target task, no explicit information about the classes or any labels are needed. UMTRA uses random sampling and augmentation to create synthetic training tasks for meta-learning phase. Labels are only needed at the final target task learning step, and they can be as little as one sample per class. On the Omniglot and Mini-Imagenet few-shot learning benchmarks, UMTRA outperforms every tested approach based on unsupervised learning of representations, while alternating for the best performance with the recent CACTUs algorithm. Compared to supervised model-agnostic meta-learning approaches, UMTRA trades off some classification accuracy for a reduction in the required labels of several orders of magnitude.

Cited by in corpus (23)

• Meta-Learning Requires Meta-Augmentation
• FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding
• Diversity Helps: Unsupervised Few-shot Learning via Distribution Shift-based Data Augmentation
• Towards Cross-Granularity Few-Shot Learning: Coarse-to-Fine Pseudo-Labeling with Visual-Semantic Meta-Embedding
• Signal Transformer: Complex-valued Attention and Meta-Learning for Signal Recognition
• Contrastive Prototype Learning with Augmented Embeddings for Few-Shot Learning
• ProtoTransformer: A Meta-Learning Approach to Providing Student Feedback
• Multiple Meta-model Quantifying for Medical Visual Question Answering
• Machine learning with limited data
• SketchEmbedNet: Learning Novel Concepts by Imitating Drawings
• Few-Shot Unsupervised Continual Learning through Meta-Examples
• Unsupervised Meta-Learning through Latent-Space Interpolation in Generative Models
• Are Fewer Labels Possible for Few-shot Learning?
• Model-Agnostic Learning to Meta-Learn
• Uniform Priors for Data-Efficient Transfer
• Trainable Class Prototypes for Few-Shot Learning
• Semi-supervised Learning with a Teacher-student Network for Generalized Attribute Prediction
• Meta-learning of Pooling Layers for Character Recognition
• Online Unsupervised Learning of Visual Representations and Categories
• Few-Shot Learning with Part Discovery and Augmentation from Unlabeled Images
• Unsupervised Few-Shot Action Recognition via Action-Appearance Aligned Meta-Adaptation
• Dual Prototypical Contrastive Learning for Few-shot Semantic Segmentation
• Divergent Search for Few-Shot Image Classification