Abstract:In Few-Shot Learning (FSL), the objective is to correctly recognize new samples from novel classes with only a few available samples per class. Existing methods in FSL primarily focus on learning transferable knowledge from base classes by maximizing the information between feature representations and their corresponding labels. However, this approach may suffer from the "supervision collapse" issue, which arises due to a bias towards the base classes. In this paper, we propose a solution to address this issue by preserving the intrinsic structure of the data and enabling the learning of a generalized model for the novel classes. Following the InfoMax principle, our approach maximizes two types of mutual information (MI): between the samples and their feature representations, and between the feature representations and their class labels. This allows us to strike a balance between discrimination (capturing class-specific information) and generalization (capturing common characteristics across different classes) in the feature representations. To achieve this, we adopt a unified framework that perturbs the feature embedding space using two low-bias estimators. The first estimator maximizes the MI between a pair of intra-class samples, while the second estimator maximizes the MI between a sample and its augmented views. This framework effectively combines knowledge distillation between class-wise pairs and enlarges the diversity in feature representations. By conducting extensive experiments on popular FSL benchmarks, our proposed approach achieves comparable performances with state-of-the-art competitors. For example, we achieved an accuracy of 69.53% on the miniImageNet dataset and 77.06% on the CIFAR-FS dataset for the 5-way 1-shot task.

Feature Learning for One-Shot Face Recognition.

One-Shot Face Recognition Based on Multiple Classifiers Training

Learning Individual-Specific Dictionaries With Fused Multiple Features For Face Recognition

Self-Attentive Networks for One-Shot Image Recognition

Learning to focus: cascaded feature matching network for few-shot image recognition

Feature Space Regularization For Person Re-Identification With One Sample

Multi-Attention Network For One Shot Learning

Balancing Feature Alignment and Uniformity for Few-Shot Classification.

Frontal-Centers Guided Face: Boosting Face Recognition by Learning Pose-Invariant Features

An Embarrassingly Simple Baseline to One-shot Learning.

Robust Face Recognition via Multimodal Deep Face Representation

Single Sample Face Recognition Via Learning Deep Supervised Autoencoders

Fuzzy Sparse Autoencoder Framework for Single Image Per Person Face Recognition.

Open-Set Face Identification on Few-Shot Gallery by Fine-Tuning

Explore Efficient Local Features from RGB-D Data for One-Shot Learning Gesture Recognition

Adaptive Local Feature Matching for Few-shot Fine-grained Image Recognition

Multi-Level Semantic Feature Augmentation for One-Shot Learning

Alignment Based Matching Networks for One-Shot Classification and Open-Set Recognition

Face Recognition: A Novel Un-Supervised Convolutional Neural Network Method

CLIP Unreasonable Potential in Single-Shot Face Recognition

Scale Attention for Learning Deep Face Representation: A Study Against Visual Scale Variation