We propose an approach for training speaker identification models in a weakly supervised manner. We concentrate on the setting where the training data consists of a set of audio recordings and the speaker annotation is provided only at the recording level. The method uses speaker diarization to find unique speakers in each recording, and i-vectors to project the speech of each speaker to a fixed-dimensional vector. A neural network is then trained to map i-vectors to speakers, using a special objective function that allows to optimize the model using recording-level speaker labels. We report experiments on two different real-world datasets. On the VoxCeleb dataset, the method provides 94.6% accuracy on a closed set speaker identification task, surpassing the baseline performance by a large margin. On an Estonian broadcast news dataset, the method provides 66% time-weighted speaker identification recall at 93% precision.
Cite as: Karu, M., Alumäe, T. (2018) Weakly Supervised Training of Speaker Identification Models . Proc. The Speaker and Language Recognition Workshop (Odyssey 2018), 24-30, doi: 10.21437/Odyssey.2018-4
@inproceedings{karu18_odyssey,
author={Martin Karu and Tanel Alumäe},
title={{Weakly Supervised Training of Speaker Identification Models }},
year=2018,
booktitle={Proc. The Speaker and Language Recognition Workshop (Odyssey 2018)},
pages={24--30},
doi={10.21437/Odyssey.2018-4}
}