Effects of Long-Term Ageing on Speaker Verification

Abstract

Abstract. The changes that occur in the human voice due to ageing have been well documented. The impact of these changes on speaker verification is less clear. In this work, we examine the effect of long-term vocal ageing on a speaker verification system. On a cohort of 13 adult speakers, using a conventional GMM-UBM system, we carry out longitudinal testing of each speaker across a time span of 30-40 years. We uncover a progressive degradation in verification score as the time span between the training and test material increases. The addition of temporal information to the features causes the rate of degradation to increase. No significant difference was found between MFCC and PLP features. Subsequent experiments show that the effect of short-term ageing (<5 years) is not significant compared with normal inter-session variability. Above this time span however, ageing has a detrimental effect on verification. Finally, we show that the age of the speaker at the time of training influences the rate at which the verification scores degrade. Our results suggest that the verification score drop-off accelerates for speakers over the age of 60. The results presented are the first of their kind to quantify the effect of long-term vocal ageing on speaker verification.

Speaker verification with long-term ageing data

“…This has been observed in [14], and a similar trend has been observed in the domain of face ageing [7]. The LLR scores of the imposters experience less change over time.…”

“…It was concluded that over a time interval of this length, variability in genuine speaker scores is more attributable to inter-session variability than to ageing. Our previous work [14] analysed longitudinal speaker data over a 30-40 year period. Using a Gaussian Mixture Model -Universal Background Model (GMM-UBM) system, the log likelihood ratio (LLR) scores of the speakers' recordings against their models were presented as a function of time span between training and testing.…”

An overview of robustness related issues in speaker recognition

“…From the score domain, researchers observed that verification scores of true speakers decreased progressively as the time span between enrollment and verification increased, while impostor scores were less affected [85,86]. Thus a stacked classifier method with an ageing-dependent decision boundary was applied to improve long-term verification accuracy.…”

Age-Related Voice Disguise and its Impact on Speaker Verification Accuracy

“…A few studies have investigated the impact of voice ageing to speaker recognition accuracy [7], as well as considering the speaker's age as a factor affecting system performance evaluation [8,9]. In [7], the impact of long-term ageing on the performance of two automatic speaker verification systems (ASV) was evaluated.…”

“…A few studies have investigated the impact of voice ageing to speaker recognition accuracy [7], as well as considering the speaker's age as a factor affecting system performance evaluation [8,9]. In [7], the impact of long-term ageing on the performance of two automatic speaker verification systems (ASV) was evaluated. It was reported that age difference between enrollment and test samples increased equal error rate (EER) in the case of male speakers from 4.61%, when the difference was less than a year, to 32.74% when the age difference range was 51-60 years.…”