Abstract:Few-shot learning, which aims to learn the concept of novel category from extremely limited labeled samples, has received intense interests in remote sensing image scene classification. Most of the existing methods inherit the philosophy of prototype learning and tackle classification as the prototype-based metric matching problem. Despite the achievement that has been obtained so far, the problems of interclass metric misalignment and intraclass variations have become two main challenges that obstacle the performance. In this article, a novel transductive learning framework with conditional metric embedding is proposed to remedy these problems. First, a conditional metric embedding mechanism is introduced to perform anisotropic embedding for each pair of the support category and query instance. This design provides the model with flexible scalability to accommodate the metric biases across classes. Second, a transductive prototype learning strategy is presented to enhance the robustness of the prototype against intraclass variations. The unlabeled query instances are transformed into pseudo instances equipped with credibility coefficients and then leveraged to calibrate the prototype estimation bias in low-data regimes. Third, a long-term consistency regularization is designed, which continuously memorizes the historical prototypes to generate additional supervision of interclass separation in the global label space. Benefiting from this design, the discriminability of the prototypes obtains obvious improvement. Finally, extensive experiments are conducted on three public benchmark remote sensing datasets. The experimental results demonstrate the validity and the superiority of the proposed method in low-shot conditions.

Few-Shot Text and Image Classification via Analogical Transfer Learning

Feature Transformation for Few-Shot Learning

A Visual Analytical Approach for Transfer Learning in Classification.

Few-Shot Incremental Learning for Label-to-Image Translation

Research on a Cross-Domain Few-Shot Adaptive Classification Algorithm Based on Knowledge Distillation Technology

Adaptive Fine-Tuning Strategy for Few-Shot Learning

Anomaly Multi-classification in Industrial Scenarios: Transferring Few-shot Learning to a New Task

Analogical Learning-Based Few-Shot Class-Incremental Learning

Multi-domain few-shot image recognition with knowledge transfer

Few-shot Object Detection Via Message Transfer Mechanism

Few-Shot Transfer Learning for Text Classification With Lightweight Word Embedding Based Models

Few-Shot Image Classification: Current Status and Research Trends

Multi-directional Knowledge Transfer for Few-Shot Learning

Semantic-Based Few-Shot Learning by Interactive Psychometric Testing

Knowledge Transduction for Cross-Domain Few-Shot Learning

Context-Transformer: Tackling Object Confusion for Few-Shot Detection

TDNet: A Novel Transductive Learning Framework with Conditional Metric Embedding for Few-shot Remote Sensing Image Scene Classification

Source Data-absent Unsupervised Domain Adaptation Through Hypothesis Transfer and Labeling Transfer

Language-Inspired Relation Transfer for Few-Shot Class-Incremental Learning

Adaptive Attribute Distribution Similarity for Few-Shot Learning

Cross Modal Few-Shot Contextual Transfer for Heterogenous Image Classification