Search

We address speaker comparison by listening in a game-like environment, hypothesized to make the task more motivating for naive listeners. We present the same 30 trials selected with the help of an x-vector speaker recognition system from VoxCeleb to a total of 150 crowdworkers recruited through Amazon's Mechanical Turk. They are divided into cohorts of 50, each using one of three alternative interface designs: (i) a traditional (nongamified) design; (ii) a gamified design with feedback on decisions, along with points, game level indications, and possibility for interface customization; (iii) another gamified design with an additional constraint of maximum of 5 'lives' consumed by wrong answers. We analyze the impact of these interface designs to listener error rates (both misses and false alarms), probability calibration, time of quitting, along with survey questionnaire. The results indicate improved performance from (i) to (ii) and (iii), particularly in terms of balancing the two types of detection errors., Accepted to Odyssey 2022 The Speaker and Language Recognition Workshop

Deep learning has brought impressive progress in the study of both automatic speaker verification (ASV) and spoofing countermeasures (CM). Although solutions are mutually dependent, they have typically evolved as standalone sub-systems whereby CM solutions are usually designed for a fixed ASV system. The work reported in this paper aims to gauge the improvements in reliability that can be gained from their closer integration. Results derived using the popular ASVspoof2019 dataset indicate that the equal error rate (EER) of a state-of-the-art ASV system degrades from 1.63% to 23.83% when the evaluation protocol is extended with spoofed trials.%subjected to spoofing attacks. However, even the straightforward integration of ASV and CM systems in the form of score-sum and deep neural network-based fusion strategies reduce the EER to 1.71% and 6.37%, respectively. The new Spoofing-Aware Speaker Verification (SASV) challenge has been formed to encourage greater attention to the integration of ASV and CM systems as well as to provide a means to benchmark different solutions., 8 pages, accepted by Odyssey 2022

As automatic speaker verification (ASV) systems are vulnerable to spoofing attacks, they are typically used in conjunction with spoofing countermeasure (CM) systems to improve security. For example, the CM can first determine whether the input is human speech, then the ASV can determine whether this speech matches the speaker's identity. The performance of such a tandem system can be measured with a tandem detection cost function (t-DCF). However, ASV and CM systems are usually trained separately, using different metrics and data, which does not optimize their combined performance. In this work, we propose to optimize the tandem system directly by creating a differentiable version of t-DCF and employing techniques from reinforcement learning. The results indicate that these approaches offer better outcomes than finetuning, with our method providing a 20% relative improvement in the t-DCF in the ASVSpoof19 dataset in a constrained setting., Published in IEEE/ACM Transactions on Audio, Speech, and Language Processing. Published version available at: https://ieeexplore.ieee.org/document/9664367

In this study, we focus on nonlinear compression methods in spectral features for speaker verification based on deep neural network. We consider different kinds of channel-dependent (CD) nonlinear compression methods optimized in a data-driven manner. Our methods are based on power nonlinearities and dynamic range compression (DRC). We also propose multi-regime (MR) design on the nonlinearities, at improving robustness. Results on VoxCeleb1 and VoxMovies data demonstrate improvements brought by proposed compression methods over both the commonly-used logarithm and their static counterparts, especially for ones based on power function. While CD generalization improves performance on VoxCeleb1, MR provides more robustness on VoxMovies, with a maximum relative equal error rate reduction of 21.6%., Comment: Accepted by ICASSP2022

Due to a constantly increasing amount of speech data that is stored in different types of databases, voice privacy has become a major concern. To respond to such concern, speech researchers have developed various methods for speaker de-identification. The state-of-the-art solutions utilize deep learning solutions which can be effective but might be unavailable or impractical to apply for, for example, under-resourced languages. Formant modification is a simpler, yet effective method for speaker de-identification which requires no training data. Still, remaining intonational patterns in formant-anonymized speech may contain speaker-dependent cues. This study introduces a novel speaker de-identification method, which, in addition to simple formant shifts, manipulates f0 trajectories based on functional data analysis. The proposed speaker de-identification method will conceal plausibly identifying pitch characteristics in a phonetically controllable manner and improve formant-based speaker de-identification up to 25%., Comment: Accepted to Speech Communication. March 2022

In this paper, we initiate the concern of enhancing the spoofing robustness of the automatic speaker verification (ASV) system, without the primary presence of a separate countermeasure module. We start from the standard ASV framework of the ASVspoof 2019 baseline and approach the problem from the back-end classifier based on probabilistic linear discriminant analysis. We employ three unsupervised domain adaptation techniques to optimize the back-end using the audio data in the training partition of the ASVspoof 2019 dataset. We demonstrate notable improvements on both logical and physical access scenarios, especially on the latter where the system is attacked by replayed audios, with a maximum of 36.1% and 5.3% relative improvement on bonafide and spoofed cases, respectively. We perform additional studies such as per-attack breakdown analysis, data composition, and integration with a countermeasure system at score-level with Gaussian back-end., Comment: Accepted by Speaker Odyssey 2022

For many decades, research in speech technologies has focused upon improving reliability. With this now meeting user expectations for a range of diverse applications, speech technology is today omni-present. As result, a focus on security and privacy has now come to the fore. Here, the research effort is in its relative infancy and progress calls for greater, multidisciplinary collaboration with security, privacy, legal and ethical experts among others. Such collaboration is now underway. To help catalyse the efforts, this paper provides a high-level overview of some related research. It targets the non-speech audience and describes the benchmarking methodology that has spearheaded progress in traditional research and which now drives recent security and privacy initiatives related to voice biometrics. We describe: the ASVspoof challenge relating to the development of spoofing countermeasures; the VoicePrivacy initiative which promotes research in anonymisation for privacy preservation., Comment: Submitted to the symposium of the ISCA Security & Privacy in Speech Communications (SPSC) special interest group

Multi-taper estimators provide low-variance power spectrum estimates that can be used in place of the windowed discrete Fourier transform (DFT) to extract speech features such as mel-frequency cepstral coefficients (MFCCs). Even if past work has reported promising automatic speaker verification (ASV) results with Gaussian mixture model-based classifiers, the performance of multi-taper MFCCs with deep ASV systems remains an open question. Instead of a static-taper design, we propose to optimize the multi-taper estimator jointly with a deep neural network trained for ASV tasks. With a maximum improvement on the SITW corpus of 25.8% in terms of equal error rate over the static-taper, our method helps preserve a balanced level of leakage and variance, providing more robustness., To appear in IEEE Signal Processing Letters

ASVspoof 2021 is the forth edition in the series of bi-annual challenges which aim to promote the study of spoofing and the design of countermeasures to protect automatic speaker verification systems from manipulation. In addition to a continued focus upon logical and physical access tasks in which there are a number of advances compared to previous editions, ASVspoof 2021 introduces a new task involving deepfake speech detection. This paper describes all three tasks, the new databases for each of them, the evaluation metrics, four challenge baselines, the evaluation platform and a summary of challenge results. Despite the introduction of channel and compression variability which compound the difficulty, results for the logical access and deepfake tasks are close to those from previous ASVspoof editions. Results for the physical access task show the difficulty in detecting attacks in real, variable physical spaces. With ASVspoof 2021 being the first edition for which participants were not provided with any matched training or development data and with this reflecting real conditions in which the nature of spoofed and deepfake speech can never be predicated with confidence, the results are extremely encouraging and demonstrate the substantial progress made in the field in recent years., Accepted to the ASVspoof 2021 Workshop

Recent years have seen growing efforts to develop spoofing countermeasures (CMs) to protect automatic speaker verification (ASV) systems from being deceived by manipulated or artificial inputs. The reliability of spoofing CMs is typically gauged using the equal error rate (EER) metric. The primitive EER fails to reflect application requirements and the impact of spoofing and CMs upon ASV and its use as a primary metric in traditional ASV research has long been abandoned in favour of risk-based approaches to assessment. This paper presents several new extensions to the tandem detection cost function (t-DCF), a recent risk-based approach to assess the reliability of spoofing CMs deployed in tandem with an ASV system. Extensions include a simplified version of the t-DCF with fewer parameters, an analysis of a special case for a fixed ASV system, simulations which give original insights into its interpretation and new analyses using the ASVspoof 2019 database. It is hoped that adoption of the t-DCF for the CM assessment will help to foster closer collaboration between the anti-spoofing and ASV research communities., Published in IEEE/ACM Transactions on Audio, Speech, and Language Processing (doi updated)

In speaker verification research, objective performance benchmarking of listeners and automatic speaker verification (ASV) systems are of key importance in understanding the limits of speaker recognition. While the adoption of common data and metrics has been instrumental to progress in ASV, there are two major shortcomings. First, the utterances lack intentional voice changes imposed by the speaker. Second, the standard evaluation metrics focus on average performance across all speakers and trials. As a result, a knowledge gap remains in how the acoustic changes impact recognition performance at the level of individual speakers. This paper addresses the limits of speaker recognition in ASV systems under voice disguise using a linear mixed effects model to analyze the impact of change in long-term statistics of selected features (formants F1-F4, the bandwidths B1-B4, F0, and speaking rate) to ASV log-likelihood ratio (LLR) score. The correlations between the proposed predictive model and the LLR scores are 0.72 for females and 0.81 for male speakers. As a whole, the difference in long-term F0 between enrollment and test utterances was found to be the individually most detrimental factor, even if the ASV system uses only spectral, rather than prosodic, features.

We propose a learnable mel-frequency cepstral coefficient (MFCC) frontend architecture for deep neural network (DNN) based automatic speaker verification. Our architecture retains the simplicity and interpretability of MFCC-based features while allowing the model to be adapted to data flexibly. In practice, we formulate data-driven versions of the four linear transforms of a standard MFCC extractor -- windowing, discrete Fourier transform (DFT), mel filterbank and discrete cosine transform (DCT). Results reported reach up to 6.7\% (VoxCeleb1) and 9.7\% (SITW) relative improvement in term of equal error rate (EER) from static MFCCs, without additional tuning effort., Comment: Accepted to ISCAS 2021

Whether it be for results summarization, or the analysis of classifier fusion, some means to compare different classifiers can often provide illuminating insight into their behaviour, (dis)similarity or complementarity. We propose a simple method to derive 2D representation from detection scores produced by an arbitrary set of binary classifiers in response to a common dataset. Based upon rank correlations, our method facilitates a visual comparison of classifiers with arbitrary scores and with close relation to receiver operating characteristic (ROC) and detection error trade-off (DET) analyses. While the approach is fully versatile and can be applied to any detection task, we demonstrate the method using scores produced by automatic speaker verification and voice anti-spoofing systems. The former are produced by a Gaussian mixture model system trained with VoxCeleb data whereas the latter stem from submissions to the ASVspoof 2019 challenge., Comment: Accepted to Interspeech 2021. Example code available at https://github.com/asvspoof-challenge/classifier-adjacency

Voice anti-spoofing aims at classifying a given utterance either as a bonafide human sample, or a spoofing attack (e.g. synthetic or replayed sample). Many anti-spoofing methods have been proposed but most of them fail to generalize across domains (corpora) -- and we do not know \emph{why}. We outline a novel interpretative framework for gauging the impact of data quality upon anti-spoofing performance. Our within- and between-domain experiments pool data from seven public corpora and three anti-spoofing methods based on Gaussian mixture and convolutive neural network models. We assess the impacts of long-term spectral information, speaker population (through x-vector speaker embeddings), signal-to-noise ratio, and selected voice quality features., Comment: INTERSPEECH 2021

Deep speaker embeddings have been demonstrated to outperform their generative counterparts, i-vectors, in recent speaker verification evaluations. To combine the benefits of high performance and generative interpretation, we investigate the use of deep embedding extractor and i-vector extractor in succession. To bundle the deep embedding extractor with an i-vector extractor, we adopt aggregation layers inspired by the Gaussian mixture model (GMM) to the embedding extractor networks. The inclusion of GMM-like layer allows the discriminatively trained network to be used as a provider of sufficient statistics for the i-vector extractor to extract what we call neural i-vectors. We compare the deep embeddings to the proposed neural i-vectors on the Speakers in the Wild (SITW) and the Speaker Recognition Evaluation (SRE) 2018 and 2019 datasets. On the core-core condition of SITW, our deep embeddings obtain performance comparative to the state-of-the-art. The neural i-vectors obtain about 50% worse performance than the deep embeddings, but on the other hand outperform the previous i-vector approaches reported in the literature by a clear margin., Comment: Accepted to Odyssey 2020: The Speaker and Language Recognition Workshop. Version 2 (bugfix)

Spectrograms - time-frequency representations of audio signals - have found widespread use in neural network-based spoofing detection. While deep models are trained on the fullband spectrum of the signal, we argue that not all frequency bands are useful for these tasks. In this paper, we systematically investigate the impact of different subbands and their importance on replay spoofing detection on two benchmark datasets: ASVspoof 2017 v2.0 and ASVspoof 2019 PA. We propose a joint subband modelling framework that employs n different sub-networks to learn subband specific features. These are later combined and passed to a classifier and the whole network weights are updated during training. Our findings on the ASVspoof 2017 dataset suggest that the most discriminative information appears to be in the first and the last 1 kHz frequency bands, and the joint model trained on these two subbands shows the best performance outperforming the baselines by a large margin. However, these findings do not generalise on the ASVspoof 2019 PA dataset. This suggests that the datasets available for training these models do not reflect real world replay conditions suggesting a need for careful design of datasets for training replay spoofing countermeasures., Accepted to the Speaker Odyssey (The Speaker and Language Recognition Workshop) 2020 conference. 8 pages

Automatic speaker verification (ASV) is one of the most natural and convenient means of biometric person recognition. Unfortunately, just like all other biometric systems, ASV is vulnerable to spoofing, also referred to as "presentation attacks." These vulnerabilities are generally unacceptable and call for spoofing countermeasures or "presentation attack detection" systems. In addition to impersonation, ASV systems are vulnerable to replay, speech synthesis, and voice conversion attacks. The ASVspoof 2019 edition is the first to consider all three spoofing attack types within a single challenge. While they originate from the same source database and same underlying protocol, they are explored in two specific use case scenarios. Spoofing attacks within a logical access (LA) scenario are generated with the latest speech synthesis and voice conversion technologies, including state-of-the-art neural acoustic and waveform model techniques. Replay spoofing attacks within a physical access (PA) scenario are generated through carefully controlled simulations that support much more revealing analysis than possible previously. Also new to the 2019 edition is the use of the tandem detection cost function metric, which reflects the impact of spoofing and countermeasures on the reliability of a fixed ASV system. This paper describes the database design, protocol, spoofing attack implementations, and baseline ASV and countermeasure results. It also describes a human assessment on spoofed data in logical access. It was demonstrated that the spoofing data in the ASVspoof 2019 database have varied degrees of perceived quality and similarity to the target speakers, including spoofed data that cannot be differentiated from bona-fide utterances even by human subjects., Accepted, Computer Speech and Language. This manuscript version is made available under the CC-BY-NC-ND 4.0. For the published version on Elsevier website, please visit https://doi.org/10.1016/j.csl.2020.101114

Modern automatic speaker verification relies largely on deep neural networks (DNNs) trained on mel-frequency cepstral coefficient (MFCC) features. While there are alternative feature extraction methods based on phase, prosody and long-term temporal operations, they have not been extensively studied with DNN-based methods. We aim to fill this gap by providing extensive re-assessment of 14 feature extractors on VoxCeleb and SITW datasets. Our findings reveal that features equipped with techniques such as spectral centroids, group delay function, and integrated noise suppression provide promising alternatives to MFCCs for deep speaker embeddings extraction. Experimental results demonstrate up to 16.3\% (VoxCeleb) and 25.1\% (SITW) relative decrease in equal error rate (EER) to the baseline., Accepted to Interspeech 2020

Security of automatic speaker verification (ASV) systems is compromised by various spoofing attacks. While many types of non-proactive attacks (and their defenses) have been studied in the past, attacker's perspective on ASV, represents a far less explored direction. It can potentially help to identify the weakest parts of ASV systems and be used to develop attacker-aware systems. We present an overview on this emerging research area by focusing on potential threats of adversarial attacks on ASV, spoofing countermeasures, or both. We conclude the study with discussion on selected attacks and leveraging from such knowledge to improve defense mechanisms against adversarial attacks.

Automatic speaker verification (ASV) vendors and corpus providers would both benefit from tools to reliably extrapolate performance metrics for large speaker populations without collecting new speakers. We address false alarm rate extrapolation under a worst-case model whereby an adversary identifies the closest impostor for a given target speaker from a large population. Our models are generative and allow sampling new speakers. The models are formulated in the ASV detection score space to facilitate analysis of arbitrary ASV systems., Accepted for publication to Interspeech 2020

In shouting, speakers use increased vocal effort to convey spoken messages over distance or above environmental noise. For automatic speaker recognition systems trained using normal speech, shouting causes a severe vocal effort mismatch between the enrollment and test hence reducing the recognition performance. In this study, two compensation methods are proposed to tackle the mismatch in a shouted versus normal speaker recognition task. These techniques are applied in the feature extraction stage of a speaker recognition system to modify the spectral envelopes of shouts to be closer to those in normal speech. The techniques modify the all-pole power spectrum of the MFCC computation chain with shouted-to-normal compensation filtering that is obtained using a GMM-based statistical mapping. In an evaluation using the state-of-the-art i-vector based recognition system, the proposed techniques provided considerable improvements in identification rates compared to the case when shouted speech spectra were not processed.

The spoofing countermeasure (CM) systems in automatic speaker verification (ASV) are not typically used in isolation of each other. These systems can be combined, for example, into a cascaded system where CM produces first a decision whether the input is synthetic or bona fide speech. In case the CM decides it is a bona fide sample, then the ASV system will consider it for speaker verification. End users of the system are not interested in the performance of the individual sub-modules, but instead are interested in the performance of the combined system. Such combination can be evaluated with tandem detection cost function (t-DCF) measure, yet the individual components are trained separately from each other using their own performance metrics. In this work we study training the ASV and CM components together for a better t-DCF measure by using reinforcement learning. We demonstrate that such training procedure indeed is able to improve the performance of the combined system, and does so with more reliable results than with the standard supervised learning techniques we compare against., Odyssey 2020 The Speaker and Language Recognition Workshop. Code available at https://github.com/Miffyli/asv-cm-reinforce

The task of speaker recognition is feasible when the speakers are co-operative or wish to be recognized. While modern automatic speaker verification (ASV) systems and some listeners are good at recognizing speakers from modal, unmodified speech, the task becomes notoriously difficult in situations of deliberate voice disguise when the speaker aims at masking his or her identity. We approach voice disguise from the perspective of acoustical and perceptual analysis using a self-collected corpus of 60 native Finnish speakers (31 female, 29 male) producing utterances in normal, intended young and intended old voice modes. The normal voices form a starting point and we are interested in studying how the two disguise modes impact the acoustical parameters and perceptual speaker similarity judgments. First, we study the effect of disguise as a relative change in fundamental frequency (F0) and formant frequencies (F1 to F4) from modal to disguised utterances. Next, we investigate whether or not speaker comparisons that are deemed easy or difficult by a modern ASV system have a similar difficulty level for the human listeners. Further, we study affecting factors from listener-related self-reported information that may explain a particular listener’s success or failure in speaker similarity assessment. Our acoustic analysis reveals a systematic increase in relative change in mean F0 for the intended young voices while for the intended old voices, the relative change is less prominent in most cases. Concerning the formants F1 through F4, 29% (for male) and 30% (for female) of the utterances did not exhibit a significant change in any formant value, while the remaining ∼ 70% of utterances had significant changes in at least one formant. Our listening panel consists of 70 listeners, 32 native and 38 non-native, who listened to 24 utterance pairs selected using rankings produced by an ASV system. The results indicate that speaker pairs categorized as easy by our ASV system were also easy for the average listener. Similarly, the listeners made more errors in the difficult trials. The listening results indicate that target (same speaker) trials were more difficult for the non-native group, while the performance for the non-target pairs was similar for both native and non-native groups., final draft, peerReviewed

International audience; In this work, we present the system description of the UIAI entry for the short-duration speaker verification (SdSV) challenge 2020. Our focus is on Task 1 dedicated to text-dependent speaker verification. We investigate different feature extraction and modeling approaches for automatic speaker verification (ASV) and utterance verification (UV). We have also studied different fusion strategies for combining UV and ASV modules. Our primary submission to the challenge is the fusion of seven subsystems which yields a normalized minimum detection cost function (minDCF) of 0.072 and an equal error rate (EER) of 2.14% on the evaluation set. The single system consisting of a pass-phrase identification based model with phone-discriminative bottleneck features gives a normalized minDCF of 0.118 and achieves 19% relative improvement over the state-of-the-art challenge baseline.

International audience; ASVspoof, now in its third edition, is a series of community-led challenges which promote the development of countermeasures to protect automatic speaker verification (ASV) from the threat of spoofing. Advances in the 2019 edition include: (i) a consideration of both logical access (LA) and physical access (PA) scenarios and the three major forms of spoofing attack, namely synthetic, converted and replayed speech; (ii) spoofing attacks generated with state-of-the-art neu-ral acoustic and waveform models; (iii) an improved, controlled simulation of replay attacks; (iv) use of the tandem detection cost function (t-DCF) that reflects the impact of both spoofing and countermeasures upon ASV reliability. Even if ASV remains the core focus, in retaining the equal error rate (EER) as a secondary metric, ASVspoof also embraces the growing importance of fake audio detection. ASVspoof 2019 attracted the participation of 63 research teams, with more than half of these reporting systems that improve upon the performance of two baseline spoofing countermeasures. This paper describes the 2019 database, protocols and challenge results. It also outlines major findings which demonstrate the real progress made in protecting against the threat of spoofing and fake audio.

The I4U consortium was established to facilitate a joint entry to NIST speaker recognition evaluations (SRE). The latest edition of such joint submission was in SRE 2018, in which the I4U submission was among the best-performing systems. SRE'18 also marks the 10-year anniversary of I4U consortium into NIST SRE series of evaluation. The primary objective of the current paper is to summarize the results and lessons learned based on the twelve subsystems and their fusion submitted to SRE'18. It is also our intention to present a shared view on the advancements, progresses, and major paradigm shifts that we have witnessed as an SRE participant in the past decade from SRE'08 to SRE'18. In this regard, we have seen, among others , a paradigm shift from supervector representation to deep speaker embedding, and a switch of research challenge from channel compensation to domain adaptation.

International audience; Welcome to this special issue on Speaker and Language Characterization which features, among other contributions, some of the most remarkable ideas presented and discussed at Odyssey 2018: the Speaker and Language Recognition Workshop, held in Les Sables d'Olonne, France, in June 2018. This issue perpetuates the series proposed by ISCA Speaker and language Characterization Special Interest Group in coordination with ISCA Speaker Odyssey workshops [1, 2, 3]. Voice is one of the most casual modalities for natural and intuitive interactions between humans as well as between humans and machines. Voice is also a central part of our identity. Voice-based solutions are currently deployed in a growing variety of applications, including person authentication through automatic speaker verification (ASV). A related technology concerns digital cloning of personal voice characteristics for text-to-speech (TTS) and voice conversion (VC). In the last years, the impressive advancements of the VC/TTS field opened the way for numerous new consumer applications. Especially, VC is offering new solutions for privacy protection. However, VC/TTS also brings the possibility of misuse of the technology in order to spoof ASV systems (for example presentation attacks implemented using voice conversion). As a direct consequence, spoofing countermeasures raises a growing interest during the past years. Moreover, voice is a central part of our identity and is also bringing other

In this work, we simulate a scenario, where a publicly available ASV system is used to enhance mimicry attacks against another closed source ASV system. In specific, ASV technology is used to perform a similarity search between the voices of recruited attackers (6) and potential target speakers (7,365) from VoxCeleb corpora to find the closest targets for each of the attackers. In addition, we consider 'median', 'furthest', and 'common' targets to serve as a reference points. Our goal is to gain insights how well similarity rankings transfer from the attacker's ASV system to the attacked ASV system, whether the attackers are able to improve their attacks by mimicking, and how the properties of the voices of attackers change due to mimicking. We address these questions through ASV experiments, listening tests, and prosodic and formant analyses. For the ASV experiments, we use i-vector technology in the attacker side, and x-vectors in the attacked side. For the listening tests, we recruit listeners through crowdsourcing. The results of the ASV experiments indicate that the speaker similarity scores transfer well from one ASV system to another. Both the ASV experiments and the listening tests reveal that the mimicry attempts do not, in general, help in bringing attacker's scores closer to the target's. A detailed analysis shows that mimicking does not improve attacks, when the natural voices of attackers and targets are similar to each other. The analysis of prosody and formants suggests that the attackers were able to considerably change their speaking rates when mimicking, but the changes in F0 and formants were modest. Overall, the results suggest that untrained impersonators do not pose a high threat towards ASV systems, but the use of ASV systems to attack other ASV systems is a potential threat., Published in Computer Speech and Language. arXiv admin note: text overlap with arXiv:1811.03790

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., (A slightly shorter version) has been submitted to IEEE ICASSP 2019

The popularization of science can often be disregarded by scientists as it may be challenging to put highly sophisticated research into words that general public can understand. This work aims to help presenting speaker recognition research to public by proposing a publicly appealing concept for showcasing recognition systems. We leverage data from YouTube and use it in a large-scale voice search web application that finds the celebrity voices that best match to the user's voice. The concept was tested in a public event as well as "in the wild" and the received feedback was mostly positive. The i-vector based speaker identification back end was found to be fast (665 ms per request) and had a high identification accuracy (93 %) for the YouTube target speakers. To help other researchers to develop the idea further, we share the source codes of the web platform used for the demo at https://github.com/bilalsoomro/speech-demo-platform., Comment: Accepted for presentation in ICASSP 2019

Speaker verification techniques neglect the short-time variation in the feature space even though it contains speaker related attributes. We propose a simple method to capture and characterize this spectral variation through the eigenstructure of the sample covariance matrix. This covariance is computed using sliding window over spectral features. The newly formulated feature vectors representing local spectral variations are used with classical and state-of-the-art speaker recognition systems. Results on multiple speaker recognition evaluation corpora reveal that eigenvectors weighted with their normalized singular values are useful in representing local covariance information. We have also shown that local variability features can be extracted using mel frequency cepstral coefficients (MFCCs) as well as using three recently developed features: frequency domain linear prediction (FDLP), mean Hilbert envelope coefficients (MHECs) and power-normalized cepstral coefficients (PNCCs). Since information conveyed in the proposed feature is complementary to the standard short-term features, we apply different fusion techniques. We observe considerable relative improvements in speaker verification accuracy in combined mode on text-independent (NIST SRE) and text-dependent (RSR2015) speech corpora. We have obtained up to 12.28% relative improvement in speaker recognition accuracy on text-independent corpora. Conversely in experiments on text-dependent corpora, we have achieved up to 40% relative reduction in EER. To sum up, combining local covariance information with the traditional cepstral features holds promise as an additional speaker cue in both text-independent and text-dependent recognition.

The papers in this special issue focus on automatic speaker verification (ASV) technologies and applications for their use. ASV offers a low-cost and flexible solution to biometric authentication. While there liability of ASV systems is now considered sufficient to support mass-market adoption, there are concerns that the technology is vulnerable to spoofing, also referred to as presentation attacks. Spoofing refers to an attack whereby a fraudster attempts to manipulate a biometric system by masquerading as another, enrolled person. Replayed, synthesized and converted speech spoofing attacks can all be used to present high-quality, convincing speech signals which are representative of other, specific speakers and thus present a genuine threat to the reliability of ASV authentication systems.

Automatic speaker verification (ASV) systems are highly vulnerable to presentation attacks, also called spoofing attacks. Replay is among the simplest attacks to mount — yet difficult to detect reliably. The generalization failure of spoofing countermeasures (CMs) has driven the community to study various alternative deep learning CMs. The majority of them are supervised approaches that learn a human-spoof discriminator. In this paper, we advocate a different, deep generative approach that leverages from powerful unsupervised manifold learning in classification. The potential benefits include the possibility to sample new data, and to obtain insights to the latent features of genuine and spoofed speech. To this end, we propose to use variational autoencoders (VAEs) as an alternative backend for replay attack detection, via three alternative models that differ in their class-conditioning. The first one, similar to the use of Gaussian mixture models (GMMs) in spoof detection, is to train independently two VAEs — one for each class. The second one is to train a single conditional model (C-VAE) by injecting a one-hot class label vector to the encoder and decoder networks. Our final proposal integrates an auxiliary classifier to guide the learning of the latent space. Our experimental results using constant-Q cepstral coefficient (CQCC) features on the ASVspoof 2017 and 2019 physical access subtask datasets indicate that the C-VAE offers substantial improvement in comparison to training two separate VAEs for each class. On the 2019 dataset, the C-VAE outperforms the VAE and the baseline GMM by an absolute 9 - 10% in both equal error rate (EER) and tandem detection cost function (t-DCF) metrics. Finally, we propose VAE residuals — the absolute difference of the original input and the reconstruction as features for spoofing detection. The proposed frontend approach augmented with a convolutional neural network classifier demonstrated substantial improvement over the VAE backend use case.

Speaker embeddings are continuous-value vector representations that allow easy comparison between voices of speakers with simple geometric operations. Among others, i-vector and x-vector have emerged as the mainstream methods for speaker embedding. In this paper, we illustrate the use of modern computation platform to harness the benefit of GPU acceleration for i-vector extraction. In particular, we achieve an acceleration of 3000 times in frame posterior computation compared to real time and 25 times in training the i-vector extractor compared to the CPU baseline from Kaldi toolkit. This significant speed-up allows the exploration of ideas that were hitherto impossible. In particular, we show that it is beneficial to update the universal background model (UBM) and re-compute frame alignments while training the i-vector extractor. Additionally, we are able to study different variations of i-vector extractors more rigorously than before. In this process, we reveal some undocumented details of Kaldi's i-vector extractor and show that it outperforms the standard formulation by a margin of 1 to 2% when tested with VoxCeleb speaker verification protocol. All of our findings are asserted by ensemble averaging the results from multiple runs with random start., Comment: Accepted to Interspeech 2019

Kinship verification is a challenging problem, where recognition systems are trained to establish a kin relation between two individuals based on facial images or videos. However, due to variations in capture conditions (background, pose, expression, illumination and occlusion), state-of-the-art systems currently provide a low level of accuracy. As in many visual recognition and affective computing applications, kinship verification may benefit from a combination of discriminant information extracted from both video and audio signals. In this paper, we investigate for the first time the fusion audio-visual information from both face and voice modalities to improve kinship verification accuracy. First, we propose a new multi-modal kinship dataset called TALking KINship (TALKIN), that is comprised of several pairs of video sequences with subjects talking. State-of-the-art conventional and deep learning models are assessed and compared for kinship verification using this dataset. Finally, we propose a deep Siamese network for multi-modal fusion of kinship relations. Experiments with the TALKIN dataset indicate that the proposed Siamese network provides a significantly higher level of accuracy over baseline uni-modal and multi-modal fusion techniques for kinship verification. Results also indicate that audio (vocal) information is complementary and useful for kinship verification problem.

Over the past few years significant progress has been made in the field of presentation attack detection (PAD) for automatic speaker recognition (ASV). This includes the development of new speech corpora, standard evaluation protocols and advancements in front-end feature extraction and back-end classifiers. The use of standard databases and evaluation protocols has enabled for the first time the meaningful benchmarking of different PAD solutions. This chapter summarises the progress, with a focus on studies completed in the last three years. The article presents a summary of findings and lessons learned from two ASVspoof challenges, the first community-led benchmarking efforts. These show that ASV PAD remains an unsolved problem and that further attention is required to develop generalised PAD solutions which have potential to detect diverse and previously unseen spoofing attacks., final draft, peerReviewed

Automatic speaker verification (ASV) systems use a playback detector to filter out playback attacks and ensure verification reliability. Since current playback detection models are almost always trained using genuine and played-back speech, it may be possible to degrade their performance by transforming the acoustic characteristics of the played-back speech close to that of the genuine speech. One way to do this is to enhance speech "stolen" from the target speaker before playback. We tested the effectiveness of a playback attack using this method by using the speech enhancement generative adversarial network to transform acoustic characteristics. Experimental results showed that use of this "enhanced stolen speech" method significantly increases the equal error rates for the baseline used in the ASVspoof 2017 challenge and for a light convolutional neural network-based method. The results also showed that its use degrades the performance of a Gaussian mixture model-universal background model-based ASV system. This type of attack is thus an urgent problem needing to be solved., Accepted at WIFS2018

The vulnerability of automatic speaker verification (ASV) systems to spoofing is widely acknowledged. Recent years have seen an intensification in research efforts to develop spoofing countermeasures, also known as presentation attack detection (PAD) systems. Much of this work has involved the exploration of features that discriminate reliably between bona fide and spoofed speech. While there are grounds to use different frontends for ASV and PAD systems (they are different tasks) the use of a single front-end has obvious benefits, not least convenience and computational efficiency, especially when ASV and PAD are combined. This paper investigates the performance of a variety of different features used previously for both ASV and PAD and assesses their performance when combined for both tasks. The paper also presents a Gaussian back-end fusion approach to system combination. In contrast to cascaded architectures, it relies upon the modelling of the two-dimensional score distribution stemming from the combination of ASV and PAD in parallel. This approach to combination is shown to generalise particularly well across independent ASVspoof 2017 v2.0 development and evaluation datasets.Index Terms: automatic speaker verification, spoofing, countermeasures, presentation attack detection

International audience; Audiovisual speech synchrony detection is an important part of talking-face verification systems. Prior work has primarily focused on visual features and joint-space models, while standard mel-frequency cepstral coefficients (MFCCs) have been commonly used to present speech. We focus more closely on audio by studying the impact of context window length for delta feature computation and comparing MFCCs with simpler energy-based features in lip-sync detection. We select state-of-the-art hand-crafted lip-sync visual features, space-time auto-correlation of gradients (STACOG), and canonical correlation analysis (CCA), for joint-space modeling. To enhance joint space modeling, we adopt deep CCA (DCCA), a nonlinear extension of CCA. Our results on the XM2VTS data indicate substantially enhanced audiovisual speech synchrony detection, with an equal error rate (EER) of 3.68%. Further analysis reveals that failed lip region localization and beard-edness of the subjects constitutes most of the errors. Thus, the lip motion description is the bottleneck, while the use of novel audio features or joint-modeling techniques is unlikely to boost lip-sync detection accuracy further.

Voice conversion (VC) aims at conversion of speaker characteristic without altering content. Due to training data limitations and modeling imperfections, it is difficult to achieve believable speaker mimicry without introducing processing artifacts; performance assessment of VC, therefore, usually involves both speaker similarity and quality evaluation by a human panel. As a time-consuming, expensive, and non-reproducible process, it hinders rapid prototyping of new VC technology. We address artifact assessment using an alternative, objective approach leveraging from prior work on spoofing countermeasures (CMs) for automatic speaker verification. Therein, CMs are used for rejecting `fake' inputs such as replayed, synthetic or converted speech but their potential for automatic speech artifact assessment remains unknown. This study serves to fill that gap. As a supplement to subjective results for the 2018 Voice Conversion Challenge (VCC'18) data, we configure a standard constant-Q cepstral coefficient CM to quantify the extent of processing artifacts. Equal error rate (EER) of the CM, a confusability index of VC samples with real human speech, serves as our artifact measure. Two clusters of VCC'18 entries are identified: low-quality ones with detectable artifacts (low EERs), and higher quality ones with less artifacts. None of the VCC'18 systems, however, is perfect: all EERs are < 30 % (the `ideal' value would be 50 %). Our preliminary findings suggest potential of CMs outside of their original application, as a supplemental optimization and benchmarking tool to enhance VC technology., Correction (bug fix) of a published ODYSSEY 2018 publication with the same title and author list; more details in footnote in page 1

We present the Voice Conversion Challenge 2018, designed as a follow up to the 2016 edition with the aim of providing a common framework for evaluating and comparing different state-of-the-art voice conversion (VC) systems. The objective of the challenge was to perform speaker conversion (i.e. transform the vocal identity) of a source speaker to a target speaker while maintaining linguistic information. As an update to the previous challenge, we considered both parallel and non-parallel data to form the Hub and Spoke tasks, respectively. A total of 23 teams from around the world submitted their systems, 11 of them additionally participated in the optional Spoke task. A large-scale crowdsourced perceptual evaluation was then carried out to rate the submitted converted speech in terms of naturalness and similarity to the target speaker identity. In this paper, we present a brief summary of the state-of-the-art techniques for VC, followed by a detailed explanation of the challenge tasks and the results that were obtained., Comment: Accepted for Speaker Odyssey 2018

International audience; The now-acknowledged vulnerabilities of automatic speaker verification (ASV) technology to spoofing attacks have spawned interests to develop so-called spoofing countermeasures. By providing common databases, protocols and metrics for their assessment, the ASVspoof initiative was born to spear-head research in this area. The first competitive ASVspoof challenge held in 2015 focused on the assessment of countermeasures to protect ASV technology from voice conversion and speech synthesis spoofing attacks. The second challenge switched focus to the consideration of replay spoofing attacks and countermeasures. This paper describes Version 2.0 of the ASVspoof 2017 database which was released to correct data anomalies detected post-evaluation. The paper contains as-yet unpublished meta-data which describes recording and playback devices and acoustic environments. These support the analysis of replay detection performance and limits. Also described are new results for the official ASVspoof baseline system which is based upon a constant Q cesptral coefficient frontend and a Gaussian mixture model backend. Reported are enhancements to the baseline system in the form of log-energy coefficients and cepstral mean and variance normalisation in addition to an alternative i-vector backend. The best results correspond to a 48% relative reduction in equal error rate when compared to the original baseline system.

The fundamental frequency (F0) contour of speech is a key aspect to represent speech prosody that finds use in speech and spoken language analysis such as voice conversion and speech synthesis as well as speaker and language identification. This work proposes new methods to estimate the F0 contour of speech using deep neural networks (DNNs) and recurrent neural networks (RNNs). They are trained using supervised learning with the ground truth of F0 contours. The latest prior research addresses this problem first as a frame-by-frame-classification problem followed by sequence tracking using deep neural network hidden Markov model (DNN-HMM) hybrid architecture. This study, however, tackles the problem as a regression problem instead, in order to obtain F0 contours with higher frequency resolution from clean and noisy speech. Experiments using PTDB-TUG corpus contaminated with additive noise (NOISEX-92) show the proposed method improves gross pitch error (GPE) by more than 25 % at signal-to-noise ratios (SNRs) between -10 dB and +10 dB as compared with one of the most noise-robust F0 trackers, PEFAC. Furthermore, the performance on fine pitch error (FPE) is improved by approximately 20 % against a state-of-the-art DNN-HMM-based approach.

The front-end factor analysis (FEFA), an extension of principal component analysis (PPCA) tailored to be used with Gaussian mixture models (GMMs), is currently the prevalent approach to extract compact utterance-level features (i-vectors) for automatic speaker verification (ASV) systems. Little research has been conducted comparing FEFA to the conventional PPCA applied to maximum a posteriori (MAP) adapted GMM supervectors. We study several alternative methods, including PPCA, factor analysis (FA), and two supervised approaches, supervised PPCA (SPPCA) and the recently proposed probabilistic partial least squares (PPLS), to compress MAP-adapted GMM supervectors. The resulting i-vectors are used in ASV tasks with a probabilistic linear discriminant analysis (PLDA) back-end. We experiment on two different datasets, on the telephone condition of NIST SRE 2010 and on the recent VoxCeleb corpus collected from YouTube videos containing celebrity interviews recorded in various acoustical and technical conditions. The results suggest that, in terms of ASV accuracy, the supervector compression approaches are on a par with FEFA. The supervised approaches did not result in improved performance. In comparison to FEFA, we obtained more than hundred-fold (100x) speedups in the total variability model (TVM) training using the PPCA and FA supervector compression approaches., To appear in Speaker Odyssey 2018: The Speaker and Language Recognition Workshop

We propose a unified approach to automatic foreign accent recognition. It takes advantage of recent technology advances in both linguistics and acoustics based modeling techniques in automatic speech recognition (ASR) while overcoming the issue of a lack of a large set of transcribed data often required in designing state-of-the-art ASR systems. The key idea lies in defining a common set of fundamental units “universally” across all spoken accents such that any given spoken utterance can be transcribed with this set of “accent-universal” units. In this study, we adopt a set of units describing manner and place of articulation as speech attributes. These units exist in most spoken languages and they can be reliably modeled and extracted to represent foreign accent cues. We also propose an i-vector representation strategy to model the feature streams formed by concatenating these units. Testing on both the Finnish national foreign language certificate (FSD) corpus and the English NIST 2008 SRE corpus, the experimental results with the proposed approach demonstrate a significant system performance improvement with p-value $ over those with the conventional spectrum-based techniques. We observed up to a 15% relative error reduction over the already very strong i-vector accented recognition system when only manner information is used. Additional improvement is obtained by adding place of articulation clues along with context information. Furthermore, diagnostic information provided by the proposed approach can be useful to the designers to further enhance the system performance.

In this work, we compare the performance of three modern speaker verification systems and non-expert human listeners in the presence of voice mimicry. Our goal is to gain insights on how vulnerable speaker verification systems are to mimicry attack and compare it to the performance of human listeners. We study both traditional Gaussian mixture model-universal background model (GMM-UBM) and an i-vector based classifier with cosine scoring and probabilistic linear discriminant analysis (PLDA) scoring. For the studied material in Finnish language, the mimicry attack decreased lightly the equal error rate (EER) for GMM-UBM from 10.83 to 10.31, while for i-vector systems the EER increased from 6.80 to 13.76 and from 4.36 to 7.38. The performance of the human listening panel shows that imitated speech increases the difficulty of the speaker verification task. It is even more difficult to recognize a person who is intentionally concealing his or her identity. For Impersonator A, the average listener made 8 errors from 34 trials while the automatic systems had 6 errors in the same set. The average listener for Impersonator B made 7 errors from the 28 trials, while the automatic systems made 7 to 9 errors. A statistical analysis of the listener performance was also conducted. We found out a statistically significant association, with p ¼ 0:00019 and R 2 ¼ 0:59, between listener accuracy and self reported factors only when familiar voices were present in the test.

Any biometric recognizer is vulnerable to spoofing attacks and hence voice biometric, also called automatic speaker verification (ASV), is no exception; replay, synthesis, and conversion attacks all provoke false acceptances unless countermeasures are used. We focus on voice conversion (VC) attacks considered as one of the most challenging for modern recognition systems. To detect spoofing, most existing countermeasures assume explicit or implicit knowledge of a particular VC system and focus on designing discriminative features. In this paper, we explore back-end generative models for more generalized countermeasures. In particular, we model synthesis-channel subspace to perform speaker verification and antispoofing jointly in the ${i}$ -vector space, which is a well-established technique for speaker modeling. It enables us to integrate speaker verification and antispoofing tasks into one system without any fusion techniques. To validate the proposed approach, we study vocoder-matched and vocoder-mismatched ASV and VC spoofing detection on the NIST 2006 speaker recognition evaluation data set. Promising results are obtained for standalone countermeasures as well as their combination with ASV systems using score fusion and joint approach.