This document provides a brief description of the National Institute of Standards and Technology (NIST) speaker recognition evaluation (SRE) conversational telephone speech (CTS) Superset. The CTS Superset has been created in an attempt to provide the research community with a large-scale dataset along with uniform metadata that can be used to effectively train and develop telephony (narrowband) speaker recognition systems. It contains a large number of telephony speech segments from more than 6800 speakers with speech durations distributed uniformly in the [10s, 60s] range. The segments have been extracted from the source corpora used to compile prior SRE datasets (SRE1996-2012), including the Greybeard corpus as well as the Switchboard and Mixer series collected by the Linguistic Data Consortium (LDC). In addition to the brief description, we also report speaker recognition results on the NIST 2020 CTS Speaker Recognition Challenge, obtained using a system trained with the CTS Superset. The results will serve as a reference baseline for the challenge.

The human voice conveys unique characteristics of an individual, making voice biometrics a key technology for verifying identities in various industries. Despite the impressive progress of speaker recognition systems in terms of accuracy, a number of ethical and legal concerns has been raised, specifically relating to the fairness of such systems. In this paper, we aim to explore the disparity in performance achieved by state-of-the-art deep speaker recognition systems, when different groups of individuals characterized by a common sensitive attribute (e.g., gender) are considered. In order to mitigate the unfairness we uncovered by means of an exploratory study, we investigate whether balancing the representation of the different groups of individuals in the training set can lead to a more equal treatment of these demographic groups. Experiments on two state-of-the-art neural architectures and a large-scale public dataset show that models trained with demographically-balanced training sets exhibit a fairer behavior on different groups, while still being accurate. Our study is expected to provide a solid basis for instilling beyond-accuracy objectives (e.g., fairness) in speaker recognition.

Speaker recognition refers to audio biometrics that utilizes acoustic characteristics for automatic speaker recognition. These systems have emerged as an essential means of verifying identity in various scenarios, such as smart homes, general business interactions, e-commerce applications, and forensics. However, the mismatch between training and real-world data causes a shift of speaker embedding space and severely degrades the recognition performance. Various complicated neural architectures are presented to address speaker recognition in the wild but neglect the requirements of storage and computation. To address this issue, we propose a neural architecture search-based efficient time-delay neural network (EfficientTDNN) to improve inference efficiency while maintaining recognition accuracy. The proposed EfficientTDNN contains three phases. First, supernet design is to construct a dynamic neural architecture that consists of sequential cells and enables network pruning. Second, progressive training is to optimize randomly sampled subnets that inherit the weights of the supernet. Third, three search methods, including manual grid search, random search, and model predictive evolutionary search, are introduced to find a trade-off between accuracy and efficiency. Results of experiments on the VoxCeleb dataset show EfficientTDNN provides a huge search space including approximately $10^{13}$ subnets and achieves 1.66% EER and 0.156 DCF$_{0.01}$ with 565M MACs. Comprehensive investigation suggests that the trained supernet generalizes cells unseen during training and obtains an acceptable balance between accuracy and efficiency.

In this article, we first establish the theory of optimal scores for speaker recognition. Our analysis shows that the minimum Bayes risk (MBR) decisions for both the speaker identification and speaker verification tasks can be based on a normalized likelihood (NL). When the underlying generative model is a linear Gaussian, the NL score is mathematically equivalent to the PLDA likelihood ratio, and the empirical scores based on cosine distance and Euclidean distance can be seen as approximations of this linear Gaussian NL score under some conditions. We discuss a number of properties of the NL score and perform a simple simulation experiment to demonstrate the properties of the NL score.

Speaker recognition is an active research area that contains notable usage in biometric security and authentication system. Currently, there exist many well-performing models in the speaker recognition domain. However, most of the advanced models implement deep learning that requires GPU support for real-time speech recognition, and it is not suitable for low-end devices. In this paper, we propose a lightweight text-independent speaker recognition model based on random forest classifier. It also introduces new features that are used for both speaker verification and identification tasks. The proposed model uses human speech based timbral properties as features that are classified using random forest. Timbre refers to the very basic properties of sound that allow listeners to discriminate among them. The prototype uses seven most actively searched timbre properties, boominess, brightness, depth, hardness, roughness, sharpness, and warmth as features of our speaker recognition model. The experiment is carried out on speaker verification and speaker identification tasks and shows the achievements and drawbacks of the proposed model. In the speaker identification phase, it achieves a maximum accuracy of 78%. On the contrary, in the speaker verification phase, the model maintains an accuracy of 80% having an equal error rate (ERR) of 0.24.

Speaker recognition is a well known and studied task in the speech processing domain. It has many applications, either for security or speaker adaptation of personal devices. In this paper, we present a new paradigm for automatic speaker recognition that we call Interactive Speaker Recognition (ISR). In this paradigm, the recognition system aims to incrementally build a representation of the speakers by requesting personalized utterances to be spoken in contrast to the standard text-dependent or text-independent schemes. To do so, we cast the speaker recognition task into a sequential decision-making problem that we solve with Reinforcement Learning. Using a standard dataset, we show that our method achieves excellent performance while using little speech signal amounts. This method could also be applied as an utterance selection mechanism for building speech synthesis systems.

Today's interactive devices such as smart-phone assistants and smart speakers often deal with short-duration speech segments. As a result, speaker recognition systems integrated into such devices will be much better suited with models capable of performing the recognition task with short-duration utterances. In this paper, a new deep neural network, UtterIdNet, capable of performing speaker recognition with short speech segments is proposed. Our proposed model utilizes a novel architecture that makes it suitable for short-segment speaker recognition through an efficiently increased use of information in short speech segments. UtterIdNet has been trained and tested on the V oxCeleb datasets, the latest benchmarks in speaker recognition. Evaluations for different segment durations show consistent and stable performance for short segments, with significant improvement over the previous models for segments of 2 seconds, 1 second, and especially sub-second durations (250 ms and 500 ms).

The performance of automatic speaker recognition systems degrades when facing distorted speech data containing additive noise and/or reverberation. Statistical uncertainty propagation has been introduced as a promising paradigm to address this challenge. So far, different uncertainty propagation methods have been proposed to compensate noise and reverberation in i-vectors in the context of speaker recognition. They have achieved promising results on small datasets such as YOHO and Wall Street Journal, but little or no improvement on the larger, highly variable NIST Speaker Recognition Evaluation (SRE) corpus. In this paper, we propose a complete uncertainty propagation method, whereby we model the effect of uncertainty both in the computation of unbiased Baum-Welch statistics and in the derivation of the posterior expectation of the i-vector. We conduct experiments on the NIST-SRE corpus mixed with real domestic noise and reverberation from the CHiME-2 corpus and preprocessed by multichannel speech enhancement. The proposed method improves the equal error rate (EER) by 4% relative compared to a conventional i-vector based speaker verification baseline. This is to be compared with previous methods which degrade performance.

ABSTRACT We consider technology-assisted mimicry attacks in the context of automatic speaker verification (ASV). We use ASV itself to select targeted speakers to be attacked by human-based mimicry. We recorded 6 naive mimics for whom we select target celebrities from VoxCeleb1 and VoxCeleb2 corpora (7,365 potential targets) using an i-vector system. The attacker attempts to mimic the selected target, with the utterances subjected to ASV tests using an independently developed x-vector system. Our main finding is negative: even if some of the attacker scores against the target speakers were slightly increased, our mimics did not succeed in spoofing the x-vector system. Interestingly, however, the relative ordering of the selected targets (closest, furthest, median) are consistent between the systems, which suggests some level of transferability between the systems.

Incremental improvements in accuracy of Convolutional Neural Networks are usually achieved through use of deeper and more complex models trained on larger datasets. However, enlarging dataset and models increases the computation and storage costs and cannot be done indefinitely. In this work, we seek to improve the identification and verification accuracy of a text-independent speaker recognition system without use of extra data or deeper and more complex models by augmenting the training and testing data, finding the optimal dimensionality of embedding space and use of more discriminative loss functions. Results of experiments on VoxCeleb dataset suggest that: (i) Simple repetition and random time-reversion of utterances can reduce prediction errors by up to 18%. (ii) Lower dimensional embeddings are more suitable for verification. (iii) Use of proposed logistic margin loss function leads to unified embeddings with state-of-the-art identification and competitive verification accuracies.