Wang Zhen, Liu Liu, Duan Yiqun, Tao Dacheng
IEEE Trans Neural Netw Learn Syst. 2023 Dec;34(12):9952-9965. doi: 10.1109/TNNLS.2022.3162747. Epub 2023 Nov 30.
Streaming label learning aims to model newly emerged labels for multilabel classification systems, which requires plenty of new label data for training. However, in changing environments, only a small amount of new label data can practically be collected. In this work, we formulate and study few-shot streaming label learning (FSLL), which models emerging new labels with only a few annotated examples by utilizing the knowledge learned from past labels. We propose a meta-learning framework, semantic inference network (SIN), which can learn and infer the semantic correlation between new labels and past labels to adapt FSLL tasks from a few examples effectively. SIN leverages label semantic representation to regularize the output space and acquires labelwise meta-knowledge based on gradient-based meta-learning. Moreover, SIN incorporates a novel label decision module with a meta-threshold loss to find the optimal confidence thresholds for each new label. Theoretically, we illustrate that the proposed semantic inference mechanism could constrain the complexity of hypotheses space to reduce the risk of overfitting and achieve better generalizability. Experimentally, extensive empirical results and ablation studies demonstrate the performance of SIN is superior to the prior state-of-the-art methods on FSLL.
流标签学习旨在为多标签分类系统中的新出现标签建模，这需要大量新标签数据进行训练。然而，在不断变化的环境中，实际上只能收集到少量新标签数据。在这项工作中，我们提出并研究了少样本流标签学习（FSLL），它通过利用从过去标签中学到的知识，仅用少量带注释的示例对新出现的标签进行建模。我们提出了一个元学习框架，即语义推理网络（SIN），它可以学习并推断新标签与过去标签之间的语义相关性，从而有效地从少量示例中适应FSLL任务。SIN利用标签语义表示来规范输出空间，并基于基于梯度的元学习获取逐标签的元知识。此外，SIN结合了一个新颖的标签决策模块和一个元阈值损失，以找到每个新标签所需的最佳置信度阈值。从理论上讲，我们证明了所提出的语义推理机制可以限制假设空间的复杂性，以降低过拟合风险并实现更好的泛化能力。在实验中，大量的实证结果和消融研究表明，SIN在FSLL上的性能优于先前的最先进方法。
