Frame-level speaker embeddings for text-independent speaker recognition and analysis of end-to-end model
This work addresses speaker recognition for applications like security or voice authentication, but it is incremental as it builds on existing CNN methods with a focus on analysis rather than major performance breakthroughs.
The authors tackled text-independent speaker recognition by proposing a CNN model that extracts frame-level speaker embeddings, finding that the model discriminates broad phonetic classes better than individual phonemes, with similar embeddings for the same speaker within phonetic classes.
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.