Three dimentional log Mel Spectrogram based Channel Attention Convolutional Recurrent Neural Network for Few-Shot Speaker Identification
VCTK and Voxceleb1 are publicly available.
URL for databases to download.
https://www.kaggle.com/datasets/showmik50/vctk-dataset
https://www.tensorflow.org/datasets/catalog/vctk
https://www.robots.ox.ac.uk/~vgg/data/voxceleb/vox1.html
https://www.tensorflow.org/datasets/catalog/voxceleb
IIT-MV daatabase description :Refer the attached paper.