The IBM Speaker Recognition System: Recent Advances and Error Analysis

TL;DR

This paper details recent improvements in IBM's speaker recognition system, including advanced feature extraction and modeling techniques, achieving state-of-the-art results on NIST 2010 SRE conditions and analyzing remaining errors.

Contribution

Introduces NDA for variability compensation, speaker-adapted features from ASR, and a DNN with extensive output units for improved speaker recognition accuracy.

Findings

Achieved best published results on NIST 2010 SRE conditions

System attained an EER of 0.59% on extended tel-tel condition (C5)

Error analysis identified issues with low-quality recordings and their impact on performance

Abstract

We present the recent advances along with an error analysis of the IBM speaker recognition system for conversational speech. Some of the key advancements that contribute to our system include: a nearest-neighbor discriminant analysis (NDA) approach (as opposed to LDA) for intersession variability compensation in the i-vector space, the application of speaker and channel-adapted features derived from an automatic speech recognition (ASR) system for speaker recognition, and the use of a DNN acoustic model with a very large number of output units (~10k senones) to compute the frame-level soft alignments required in the i-vector estimation process. We evaluate these techniques on the NIST 2010 SRE extended core conditions (C1-C9), as well as the 10sec-10sec condition. To our knowledge, results achieved by our system represent the best performances published to date on these conditions. For example, on the extended tel-tel condition (C5) the system achieves an EER of 0.59%. To garner further understanding of the remaining errors (on C5), we examine the recordings associated with the low scoring target trials, where various issues are identified for the problematic recordings/trials. Interestingly, it is observed that correcting the pathological recordings not only improves the scores for the target trials but also for the nontarget trials.