Most of the current excellent models in speaker verification are ResNet-based deep models and attention-based models. These models have a general weakness, which is the large number of parameters and ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results