Few-Shot Learning with Adaptive Weight Masking in Conditional GANs

TL;DR

This paper presents RWM-CGAN, a novel generative model that enhances few-shot learning by improving data augmentation quality and diversity through residual units and weight mask regularization, leading to better classification performance.

Contribution

It introduces a Residual Weight Masking Conditional GAN that effectively augments data in few-shot learning, addressing overfitting and generalization issues with a novel network architecture and regularization technique.

Findings

RWM-CGAN improves sample diversity and quality.

Significant accuracy gains in detection and classification.

Effective in small-sample category learning.

Abstract

Deep learning has revolutionized various fields, yet its efficacy is hindered by overfitting and the requirement of extensive annotated data, particularly in few-shot learning scenarios where limited samples are available. This paper introduces a novel approach to few-shot learning by employing a Residual Weight Masking Conditional Generative Adversarial Network (RWM-CGAN) for data augmentation. The proposed model integrates residual units within the generator to enhance network depth and sample quality, coupled with a weight mask regularization technique in the discriminator to improve feature learning from small-sample categories. This method addresses the core issues of robustness and generalization in few-shot learning by providing a controlled and clear augmentation of the sample space. Extensive experiments demonstrate that RWM-CGAN not only expands the sample space effectively but also enriches the diversity and quality of generated samples, leading to significant improvements in detection and classification accuracy on public datasets. The paper contributes to the advancement of few-shot learning by offering a practical solution to the challenges posed by data scarcity and the need for rapid generalization to new tasks or categories.