NeuralPlda - v1

Pytorch Implementation of the Pairwise Discriminative Neural PLDA (NPLDA) for Speaker Verification

Refs: [1] Shreyas Ramoji, Prashant Krishnan, Prachi Singh, Sriram Ganapathy: "Pairwise Discriminative Neural PLDA for Speaker Verification" arXiv preprint arXiv:2001.07034 (2020).

[2] Shreyas Ramoji, Prashant Krishnan, Sriram Ganapathy: "NPLDA: A Deep Neural PLDA Model for Speaker Verification" submitted to Odyssey 2020, The Speaker and Language Recognition Workshop.

[3] Shreyas Ramoji, Prashant Krishnan, Bhargavram Mysore, Prachi Singh, Sriram Ganapathy: "LEAP System for SRE19 Challenge - Improvements and Error Analysis" arXiv preprint arXiv:2001.07034 (2020).

Requirements:

1. Kaldi Toolkit : https://github.com/kaldi-asr/kaldi
2. Python 3.6 or above with Numpy, Scipy, Pytorch, Matplotlib, Pickle, kaldi_io (https://pypi.org/project/kaldi-io/, https://github.com/vesis84/kaldi-io-for-python)