研究人员推出了一种新颖的$\alpha$-散度损失函数Q-Margin，旨在改进生物识别系统。与将几何惩罚应用于logits的传统方法不同，这种新损失函数将一个原则性的概率边距直接编码到先验概率中。Q-Margin旨在鼓励区分性嵌入同时保持稀疏性，从而在人脸和说话人识别基准测试的低错误接受率下提高性能。 AI
影响 这项研究通过提高关键的低错误接受率下的准确性，可能带来更安全、更高效的生物识别系统。
排序理由 该集群包含一篇详细介绍生物识别新颖损失函数的论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2606.31664v1 Announce Type: cross Abstract: Performance in face and speaker verification is largely driven by margin-penalty softmax losses such as CosFace and ArcFace. Recently introduced $\alpha$-divergence loss functions offer a compelling alternative, particularly due t…
Performance in face and speaker verification is largely driven by margin-penalty softmax losses such as CosFace and ArcFace. Recently introduced $α$-divergence loss functions offer a compelling alternative, particularly due to their ability to induce sparse solutions (when $α>1$)…