Abstract: Currently, most speaker recognition systems make a decision by calculating the similarity between enrollment and test embeddings extracted with convolutional neural networks. However, for ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback