Frame-level speaker embeddings for text-independent speaker recognition and analysis of end-to-end model

TL;DR

This paper introduces a CNN-based speaker recognition model that extracts frame-level embeddings, enabling detailed analysis of speech features and improving understanding of how the model differentiates voice identities in a text-independent setting.

Contribution

The paper presents a novel CNN architecture capable of extracting frame-level speaker embeddings and analyzes its ability to distinguish phonetic classes and speaker identities.

Findings

Networks better discriminate broad phonetic classes than individual phonemes.

Frame-level embeddings are similar for the same phonetic class within a speaker.

Potential for improved speaker recognition through frame-level analysis.

Abstract

In this paper, we propose a Convolutional Neural Network (CNN) based speaker recognition model for extracting robust speaker embeddings. The embedding can be extracted efficiently with linear activation in the embedding layer. To understand how the speaker recognition model operates with text-independent input, we modify the structure to extract frame-level speaker embeddings from each hidden layer. We feed utterances from the TIMIT dataset to the trained network and use several proxy tasks to study the networks ability to represent speech input and differentiate voice identity. We found that the networks are better at discriminating broad phonetic classes than individual phonemes. In particular, frame-level embeddings that belong to the same phonetic classes are similar (based on cosine distance) for the same speaker. The frame level representation also allows us to analyze the networks at the frame level, and has the potential for other analyses to improve speaker recognition.