Deep Speaker Embeddings for Short-Duration Speaker Verification

Abstract

The performance of a state-of-the-art speaker verification system is severely degraded when it is presented with trial recordings of short duration. In this work we propose to use deep neural networks to learn short-duration speaker embeddings. We focus on the 5s-5s condition, wherein both sides of a verification trial are 5 seconds long. In our previous work we established that learning a non-linear mapping from i-vectors to speaker labels is beneficial for speaker verification [1]. In this work we take the idea of learning a speaker classifier one step further-we apply deep neural networks directly to timefrequency speech representations. We propose two feedforward network architectures for this task. Our best model is based on a deep convolutional architecture wherein recordings are treated as images. From our experimental findings we advocate treating utterances as images or 'speaker snapshots, much like in face recognition. Our convolutional speaker embeddings perform significantly better than i-vectors when scoring is done using cosine distance, where the relative improvement is 23.5%. The proposed deep embeddings combined with cosine distance also outperform a state-of-the-art i-vector verification system by 1%, providing further empirical evidence in favor of our learned speaker features.

Within-Sample Variability-Invariant Loss for Robust Speaker Recognition Under Noisy Environments

“…Self-attentive pooling layer [20], learnable dictionary encoding layer [21], and dictionary-based NetVLAD layer [22,23] are other commonly used encoding layers. Once the utterance-level representation is extracted, a fully connected layer and a speaker classifier are employed to further abstract the speaker representation and classify the training speakers.…”

An Improved Deep Neural Network for Modeling Speaker Characteristics at Different Temporal Scales

“…h r hr (5) where is the elementwise multiplication, and D  r is the transformed i-vector through a non-linearly affine layer. Since each utterance owns an utterance-dependent i-vector, the attention weights are determined by both t h and the utterance-level information from the i-vector.…”

“…With the great success of deep learning over a wide range of machine learning tasks, more efforts have been focused on using deep neural network (DNN) to extract more discriminative speaker representations [3,4,5,6]. These deep speaker embedding systems can achieve comparable or even better performance compared with the i-vector based methods, particularly under conditions of short-duration utterances.…”

Partial AUC Optimization Based Deep Speaker Embeddings with Class-Center Learning for Text-Independent Speaker Verification

“…The state-of-the-art text-independent speaker verification systems [1][2][3][4] use deep neural networks (DNNs) to project speech recordings with different lengths into a common low dimensional embedding space where the speakers' identities are represented. Such a method is called deep embedding, where the embedding networks have three key components-network structure [1,3,[5][6][7], pooling layer [1,[8][9][10][11][12], and loss function [13][14][15][16][17]. This paper focuses on the last part, i.e., the loss functions.…”