We leveraged 3D convolutional architecture for creating the speaker model in order to simultaneously capturing the speech-related and temporal information from

We leveraged 3D convolutional architecture for creating the speaker model in order to simultaneously capturing the speech-related and temporal information from the speakers' utterances.