论文标题
Deep Speaker嵌入的神经判别分析
Neural Discriminant Analysis for Deep Speaker Embedding
论文作者
论文摘要
概率线性判别分析（PLDA）是开放式分类/验证任务中的流行工具。但是，高斯基础PLDA的假设使其无法应用于显然是非高斯的情况。在本文中，我们提出了一种新型的非线性版本的PLDA，称为神经判别分析（NDA）。该模型采用可逆的深神经网络将复杂的分布转换为简单的高斯，因此可以在变换的空间中很容易建立线性高斯模型。我们在说话者识别任务上测试了这种NDA模型，其中深层扬声器矢量（X-矢量）大概是非高斯的。两个数据集上的实验结果表明，通过处理X-VECTOR的非高斯分布，NDA始终优于PLDA。
Probabilistic Linear Discriminant Analysis (PLDA) is a popular tool in open-set classification/verification tasks. However, the Gaussian assumption underlying PLDA prevents it from being applied to situations where the data is clearly non-Gaussian. In this paper, we present a novel nonlinear version of PLDA named as Neural Discriminant Analysis (NDA). This model employs an invertible deep neural network to transform a complex distribution to a simple Gaussian, so that the linear Gaussian model can be readily established in the transformed space. We tested this NDA model on a speaker recognition task where the deep speaker vectors (x-vectors) are presumably non-Gaussian. Experimental results on two datasets demonstrate that NDA consistently outperforms PLDA, by handling the non-Gaussian distributions of the x-vectors.
