Unsupervised Clustering Approaches for Domain Adaptation in Speaker Recognition Systems

In this paper, we motivate and define the domain adaptation challenge task for speaker recognition. Using an i-vector sys-tem trained only on out-of-domain data as a starting point, we propose a framework that utilizes large-scale clustering algo-rithms and unlabeled in-domain data to adapt the system for evaluation. In presenting the results and analyses of an empiri-cal exploration of this problem, our initial findings suggest that, while perfect clustering yields the best results, imperfect clus-tering can still provide recognition performance within 15 % of the optimal. We further present a system that achieves recogni-tion performance comparable to one that is provided all knowl-edge of the domain mismatch, and lastly, we outline throughout this paper some of the many directions for future work that this new task provides. 1