论文标题
高斯说话者嵌入学习独立的说话者验证的学习
Gaussian speaker embedding learning for text-independent speaker verification
论文作者
论文摘要
使用深神经网络的X矢量映射任意持续时间段至固定维度的向量。结合概率线性判别分析（PLDA）后端，X-Vector/PLDA已成为独立于文本的说话者验证的主要框架。然而，如何提取适合PLDA后端的X向量是一个关键问题。在本文中，我们提出了一个高斯噪声限制网络（GNCN）来提取XVECTOR，该网络采用了多任务学习策略，主要任务将说话者分类，而辅助任务则适合高斯噪音。使用SITW数据库进行实验。结果证明了我们提出的方法的有效性
The x-vector maps segments of arbitrary duration to vectors of fixed dimension using deep neural network. Combined with the probabilistic linear discriminant analysis (PLDA) backend, the x-vector/PLDA has become the dominant framework in text-independent speaker verification. Nevertheless, how to extract the x-vector appropriate for the PLDA backend is a key problem. In this paper, we propose a Gaussian noise constrained network (GNCN) to extract xvector, which adopts a multi-task learning strategy with the primary task classifying the speakers and the auxiliary task just fitting the Gaussian noises. Experiments are carried out using the SITW database. The results demonstrate the effectiveness of our proposed method
