TapNet: Neural Network Augmented with Task-Adaptive Projection for Few-Shot Learning

Abstract

Handling previously unseen tasks after given only a few training examples continues to be a tough challenge in machine learning. We propose TapNets, neural networks augmented with task-adaptive projection for improved few-shot learning. Here, employing a meta-learning strategy with episode-based training, a network and a set of per-class reference vectors are learned across widely varying tasks. At the same time, for every episode, features in the embedding space are linearly projected into a new space as a form of quick task-specific conditioning. The training loss is obtained based on a distance metric between the query and the reference vectors in the projection space. Excellent generalization results in this way. When tested on the Omniglot, miniImageNet and tieredImageNet datasets, we obtain state of the art classification accuracies under various few-shot scenarios.

in proceedings of the 36th International Conference on Machine Learning (ICML), Long Beach, PMLR 97:7115-7123, 2019

Cited by in corpus (26)

• OSLNet: Deep Small-Sample Classification with an Orthogonal Softmax Layer
• Rectifying the Shortcut Learning of Background for Few-Shot Learning
• MHFC: Multi-Head Feature Collaboration for Few-Shot Learning
• Cross-Domain Few-Shot Learning by Representation Fusion
• Automated Relational Meta-learning
• Learning to Learn Variational Semantic Memory
• RelationNet2: Deep Comparison Columns for Few-Shot Learning
• On Episodes, Prototypical Networks, and Few-shot Learning
• Learning Adaptive Classifiers Synthesis for Generalized Few-Shot Learning
• Meta-Learning across Meta-Tasks for Few-Shot Learning
• Boosting Few-Shot Learning With Adaptive Margin Loss
• Contrastive Prototype Learning with Augmented Embeddings for Few-Shot Learning
• Online Structured Meta-learning
• XtarNet: Learning to Extract Task-Adaptive Representation for Incremental Few-Shot Learning
• ECKPN: Explicit Class Knowledge Propagation Network for Transductive Few-shot Learning
• On the Importance of Firth Bias Reduction in Few-Shot Classification
• Ranking Distance Calibration for Cross-Domain Few-Shot Learning
• Subspace Representation Learning for Few-shot Image Classification
• Few-Shot Learning with Intra-Class Knowledge Transfer
• How to trust unlabeled data? Instance Credibility Inference for Few-Shot Learning
• Task-Adaptive Feature Transformer for Few-Shot Segmentation
• Few-shot Intent Classification and Slot Filling with Retrieved Examples
• A Strong Baseline for Semi-Supervised Incremental Few-Shot Learning
• Prior Omission of Dissimilar Source Domain(s) for Cost-Effective Few-Shot Learning
• Uncertainty-Aware Few-Shot Image Classification
• MDFM: Multi-Decision Fusing Model for Few-Shot Learning