用生成模型学习类别嵌入残差，提升单样本增量学习准确率
Feature-Space Generative Models for One-Shot Class-Incremental Learning
- 通过减去类别原型构建残差空间，捕捉基类与新类的结构相似性
- 在多个基准上实现优于当前最优的单样本新类别识别性能
- 适合需要低资源增量学习的应用场景，如医疗图像分类
少样本类别增量学习（FSCIL）要求模型在初始训练后，仅凭少量数据（1样本/类）识别新增类别，且不可再训练或修改模型。本文提出一种新方法，基于基类与新类嵌入具有结构相似性的假设，将输入样本的嵌入减去其类别原型（即平均嵌入），得到残差空间。利用变分自编码器（VAE）或扩散模型对基类残差的多模态分布进行建模，将其作为结构先验，增强对新类别的识别能力。所提方法Gen1S在多个基准和主干网络上均显著优于现有方法。
原文摘要 · Abstract (English)
Few-shot class-incremental learning (FSCIL) is a paradigm where a model, initially trained on a dataset of base classes, must adapt to an expanding problem space by recognizing novel classes with limited data. We focus on the challenging FSCIL setup where a model receives only a single sample (1-shot) for each novel class and no further training or model alterations are allowed after the base training phase. This makes generalization to novel classes particularly difficult. We propose a novel approach predicated on the hypothesis that base and novel class embeddings have structural similarity. We map the original embedding space into a residual space by subtracting the class prototype (i.e., the average class embedding) of input samples. Then, we leverage generative modeling with VAE or diffusion models to learn the multi-modal distribution of residuals over the base classes, and we use this as a valuable structural prior to improve recognition of novel classes. Our approach, Gen1S, consistently improves novel class recognition over the state of the art across multiple benchmarks and backbone architectures.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。