Search

Your search keyword '"Ville Vestman"' showing total 34 results
Start Over Author "Ville Vestman"
34 results on '"Ville Vestman"'

1. ASVtorch toolkit: Speaker verification with deep neural networks

Author

Kong Aik Lee, Ville Vestman, and Tomi Kinnunen

Subjects
Speaker recognition, PyTorch, Deep learning, Computer software, QA76.75-76.765
Abstract

The human voice differs substantially between individuals. This facilitates automatic speaker verification (ASV) — recognizing a person from his/her voice. ASV accuracy has substantially increased throughout the past decade due to recent advances in machine learning, particularly deep learning methods. An unfortunate downside has been substantially increased complexity of ASV systems. To help non-experts to kick-start reproducible ASV development, a state-of-the-art toolkit implementing various ASV pipelines and functionalities is required. To this end, we introduce a new open-source toolkit, ASVtorch, implemented in Python using the widely used PyTorch machine learning framework.

Published
2021
Full Text
View/download PDF

15. The I4U Mega Fusion and Collaboration for NIST Speaker Recognition Evaluation 2016.

Author

Kong-Aik Lee, Ville Hautamäki, Tomi Kinnunen, Anthony Larcher, Chunlei Zhang, Andreas Nautsch, Themos Stafylakis, Gang Liu 0001, Mickaël Rouvier, Wei Rao, Federico Alegre, J. Ma, Man-Wai Mak, Achintya Kumar Sarkar, Héctor Delgado, Rahim Saeidi, Hagai Aronowitz, Aleksandr Sizov, Hanwu Sun, Trung Hieu Nguyen 0001, Guangsen Wang, Bin Ma 0001, Ville Vestman, Md. Sahidullah, M. Halonen, Anssi Kanervisto, Gaël Le Lan, Fahimeh Bahmaninezhad, Sergey Isadskiy, Christian Rathgeb, Christoph Busch 0001, Georgios Tzimiropoulos, Q. Qian, Z. Wang, Q. Zhao, T. Wang, H. Li, J. Xue, S. Zhu, R. Jin, T. Zhao, Pierre-Michel Bousquet, Moez Ajili, Waad Ben Kheder, Driss Matrouf, Zhi Hao Lim, Chenglin Xu, Haihua Xu, Xiong Xiao, Eng Siong Chng, Benoit G. B. Fauve, Kaavya Sriskandaraja, Vidhyasaharan Sethu, W. W. Lin, Dennis Alexander Lehmann Thomsen, Zheng-Hua Tan, Massimiliano Todisco, Nicholas W. D. Evans, Haizhou Li 0001, John H. L. Hansen, Jean-François Bonastre, and Eliathamby Ambikairajah

Published
2017
Full Text
View/download PDF

21. The ASVspoof 2019 database.

Author

Xin Wang 0037, Junichi Yamagishi, Massimiliano Todisco, Héctor Delgado, Andreas Nautsch, Nicholas W. D. Evans, Md. Sahidullah, Ville Vestman, Tomi Kinnunen, Kong Aik Lee, Lauri Juvela, Paavo Alku, Yu-Huai Peng, Hsin-Te Hwang, Yu Tsao 0001, Hsin-Min Wang, Sébastien Le Maguer, Markus Becker, Fergus Henderson, Rob Clark, Yu Zhang 0033, Quan Wang, Ye Jia, Kai Onuma, Koji Mushika, Takashi Kaneda, Yuan Jiang 0006, Li-Juan Liu, Yi-Chiao Wu, Wen-Chin Huang, Tomoki Toda, Kou Tanaka, Hirokazu Kameoka, Ingmar Steiner, Driss Matrouf, Jean-François Bonastre, Avashna Govender, Srikanth Ronanki, Jing-Xuan Zhang, and Zhen-Hua Ling

Published
2019

26. Tandem Assessment of Spoofing Countermeasures and Automatic Speaker Verification: Fundamentals

Author

Kong Aik Lee, Massimiliano Todisco, Héctor Delgado, Xin Wang, Tomi Kinnunen, Ville Vestman, Nicholas Evans, Douglas A. Reynolds, Sahidullah, Junichi Yamagishi, Andreas Nautsch, University of Eastern Finland, Eurecom [Sophia Antipolis], NEC Corporation, National Institute of Informatics (NII), Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), MIT Lincoln Laboratory, Massachusetts Institute of Technology (MIT), This work has been sponsored by Academy of Finland (proj. no. 309629), Japan Science and Technology (JST), and the Department of Defense under Air Force Contract FA8721-05-C-0002. Opinions, interpretations, conclusions and recommendations are those of the authors and are not necessarily endorsed by the United States Government. The work has also been partially funded by the EU H2020 research and innovation programme under the MSCA grant agreement No. 860813 (TReSPAsS-ETN), the ANR-DFG FrenchGerman joint project ANR-18-CE92-0024 (RESPECT), the ANR project ANR-19-DATA-0008 (HARPOCRATES), the ANR project ExTENSoR and the JST-ANR Japanese-French project VoicePersonae., ANR-18-CE92-0024,RESPECT,Authentification multi-biométrique des personnes, fiable, sécurisée et préservant la vie privée(2018), and ANR-19-DATA-0008,Harpocrates,Open data, outils et challenges pour l'anonymisation des voix(2019)

Subjects
Signal Processing (eess.SP), FOS: Computer and information sciences, Sound (cs.SD), Computer Science - Machine Learning, Spoofing attack, spoofing countermeasures, Acoustics and Ultrasonics, Computer science, media_common.quotation_subject, Word error rate, Machine learning, computer.software_genre, Computer Science - Sound, Electronic mail, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], Machine Learning (cs.LG), 030507 speech-language pathology & audiology, 03 medical and health sciences, presentation attack detection, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, [STAT.ML]Statistics [stat]/Machine Learning [stat.ML], [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Audio and Speech Processing (eess.AS), FOS: Electrical engineering, electronic engineering, information engineering, Computer Science (miscellaneous), Electrical Engineering and Systems Science - Signal Processing, Electrical and Electronic Engineering, Special case, Function (engineering), Reliability (statistics), media_common, [SPI.ACOU]Engineering Sciences [physics]/Acoustics [physics.class-ph], [STAT.AP]Statistics [stat]/Applications [stat.AP], business.industry, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], detection cost function, Speech processing, Computational Mathematics, Metric (unit), Artificial intelligence, 0305 other medical science, business, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, computer, automatic speaker verification, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

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)

Published
2020
Full Text
View/download PDF

27. Neural i-vectors

Author

Kong Aik Lee, Ville Vestman, and Tomi Kinnunen

Subjects
FOS: Computer and information sciences, Computer Science - Machine Learning, 0303 health sciences, business.industry, Computer science, Pattern recognition, Mixture model, Speaker recognition, 01 natural sciences, Machine Learning (cs.LG), Extractor, 03 medical and health sciences, Audio and Speech Processing (eess.AS), Bundle, 0103 physical sciences, FOS: Electrical engineering, electronic engineering, information engineering, Embedding, Artificial intelligence, Layer (object-oriented design), 010306 general physics, business, Sufficient statistic, Generative grammar, Electrical Engineering and Systems Science - Audio and Speech Processing, 030304 developmental biology
Abstract

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)

Published
2020
Full Text
View/download PDF

28. ASVspoof 2019: A large-scale public database of synthesized, converted and replayed speech

Author

Hirokazu Kameoka, Hsin-Te Hwang, Driss Matrouf, Markus Becker, Quan Wang, Sahidullah, Ye Jia, Yu Zhang, Lauri Juvela, Hsin-Min Wang, Wen-Chin Huang, Zhen-Hua Ling, Yuan Jiang, Yi-Chiao Wu, Héctor Delgado, Massimiliano Todisco, Yu Tsao, Li-Juan Liu, Junichi Yamagishi, Jean-François Bonastre, Tomoki Toda, Nicholas Evans, Robert A. J. Clark, Kai Onuma, Yu-Huai Peng, Sébastien Le Maguer, Avashna Govender, Takashi Kaneda, Andreas Nautsch, Kong Aik Lee, Xin Wang, Srikanth Ronanki, Ville Vestman, Koji Mushika, Ingmar Steiner, Tomi Kinnunen, Fergus Henderson, Jing-Xuan Zhang, Kou Tanaka, Paavo Alku, Hitotsubashi University, University of Edinburgh, Eurecom [Sophia Antipolis], Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), University of Eastern Finland, NEC Corporation, Aalto University, Academia Sinica, ADAPT Centre, Sigmedia Lab, EE Engineering, Trinity College Dublin, Google Inc [Mountain View], Research at Google, Hoya Corp., iFlytek Research, Nagoya City University [Nagoya, Japan], NTT Communication Science Laboratories, NTT Corporation, audEERING GmbH, Laboratoire Informatique d'Avignon (LIA), Avignon Université (AU)-Centre d'Enseignement et de Recherche en Informatique - CERI, The Centre for Speech Technology Research [Edinburgh] (CSTR), Southern University of Science and Technology (SUSTech), The work was partially supported by JST CREST Grant No. JPMJCR18A6 (VoicePersonae project), Japan, MEXT KAKENHI Grant Nos. (16H06302, 16K16096, 17H04687, 18H04120, 18H04112, 18KT0051), Japan, the VoicePersonae and RESPECT projects funded by the French Agence Nationale de la Recherche (ANR), the Academy of Finland (NOTCH project no. 309629), and Region Grand Est, France. entitled 'NOTCH: NOn-cooperaTive speaker CHaracterization'). The authors at the University of Eastern Finland also gratefully acknowledge the use of the computational infrastructures at CSC – the IT Center for Science, and the support of the NVIDIA Corporation the donation of a Titan V GPU used in this research. The numerical calculations of some of the spoofed data were carried out on the TSUBAME3.0 supercomputer at the Tokyo Institute of Technology. The work is also partially supported by Region Grand Est, France. The ADAPT centre (13/RC/2106) is funded by the Science Foundation Ireland (SFI)., National Institute of Informatics, EURECOM, Université de Lorraine, Dept Signal Process and Acoust, Trinity College Dublin, Google, USA, HOYA Corporation, IFLYTEK Co., Ltd., Nagoya University, AudEERING GmbH, Avignon Université, University of Science and Technology of China, Aalto-yliopisto, and Southern University of Science and Technology of China (SUSTech)

Subjects
Signal Processing (eess.SP), FOS: Computer and information sciences, Sound (cs.SD), ASVspoof challenge, biometrics, Computer Science - Cryptography and Security, voice conversion, Computer science, Speech synthesis, 02 engineering and technology, computer.software_genre, 01 natural sciences, Computer Science - Sound, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], text-to-speech synthesis, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Audio and Speech Processing (eess.AS), 0202 electrical engineering, electronic engineering, information engineering, Replay, Use case, media forensics, 010301 acoustics, Protocol (object-oriented programming), Text-to-speech synthesis, Database, presentation attack, ComputingMilieux_MANAGEMENTOFCOMPUTINGANDINFORMATIONSYSTEMS, Automatic speaker verification, Cryptography and Security (cs.CR), [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Electrical Engineering and Systems Science - Audio and Speech Processing, automatic speaker verification, Voice conversion, Spoofing attack, Biometrics, anti-spoofing, Reliability (computer networking), Database design, Theoretical Computer Science, replay, presentation attack detection, 0103 physical sciences, FOS: Electrical engineering, electronic engineering, information engineering, Electrical Engineering and Systems Science - Signal Processing, [SPI.ACOU]Engineering Sciences [physics]/Acoustics [physics.class-ph], ComputerSystemsOrganization_COMPUTER-COMMUNICATIONNETWORKS, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], 020206 networking & telecommunications, Human-Computer Interaction, Physical access, computer, countermeasure, Software
Abstract

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

Published
2020
Full Text
View/download PDF

29. UIAI System for Short-Duration Speaker Verification Challenge 2020

Author

Achintya Kumar Sarkar, Romain Serizel, Zheng-Hua Tan, Ville Vestman, Emmanuel Vincent, Sahidullah, Xuechen Liu, Tomi Kinnunen, Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Indian Institute of Information Technology, Sri City (IIIT Sri City), University of Eastern Finland, Aalborg University [Denmark] (AAU), IEEE, Short-duration Speaker Verification Challenge 2020, Grid'5000, Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL), and This work has been partially sponsored by Academy of Finland (proj. n° 309629).

Subjects
FOS: Computer and information sciences, Sound (cs.SD), Computer science, Speech recognition, Computer Vision and Pattern Recognition (cs.CV), Feature extraction, Computer Science - Computer Vision and Pattern Recognition, Word error rate, 02 engineering and technology, SdSV challenge 2020, X-vector, Bottleneck, Computer Science - Sound, Task (project management), Utterance verification, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], Set (abstract data type), 030507 speech-language pathology & audiology, 03 medical and health sciences, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Audio and Speech Processing (eess.AS), 0202 electrical engineering, electronic engineering, information engineering, FOS: Electrical engineering, electronic engineering, information engineering, Text-dependent speaker verification, Fusion, System fusion, [SPI.ACOU]Engineering Sciences [physics]/Acoustics [physics.class-ph], Time delay neural network, 020206 networking & telecommunications, Identification (information), Task analysis, Bottleneck feature, Time-delay neural network, Gaussian mixture model GMM, 0305 other medical science, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

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.

Published
2020
Full Text
View/download PDF

30. I4U Submission to NIST SRE 2018: Leveraging from a Decade of Shared Experiences

Author

Ville Hautamäki, Hitoshi Yamamoto, Koji Okabe, Qing Wang, Rohan Kumar Das, Hanwu Sun, Massimiliano Todisco, Chunlei Zhang, Pierre-Michel Bousquet, Kong Aik Lee, Jing Huang, Tomi Kinnunen, Ville Vestman, Wei Rao, Haizhou Li, Héctor Delgado, Fahimeh Bahmaninezhad, Guohong Ding, Mickael Rouvier, and Anthony Larcher

Subjects
Computer science, 0202 electrical engineering, electronic engineering, information engineering, NIST, Embedding, 020206 networking & telecommunications, 020201 artificial intelligence & image processing, 02 engineering and technology, Speaker recognition, Data science
Abstract

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.

Published
2019
Full Text
View/download PDF

31. Voice Mimicry Attacks Assisted by Automatic Speaker Verification

Author

Rosa González Hautamäki, Sahidullah, Tomi Kinnunen, Ville Vestman, University of Eastern Finland, Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), and Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)

Subjects
FOS: Computer and information sciences, Sound (cs.SD), Computer Science - Machine Learning, Speaker verification, Spoofing attack, Computer Science - Cryptography and Security, Computer science, Speech recognition, Nearest neighbor search, 02 engineering and technology, spoofing, Crowdsourcing, 01 natural sciences, Computer Science - Sound, Theoretical Computer Science, Machine Learning (cs.LG), [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], prosody, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Audio and Speech Processing (eess.AS), 0103 physical sciences, 0202 electrical engineering, electronic engineering, information engineering, FOS: Electrical engineering, electronic engineering, information engineering, perceptual speaker similarity, Active listening, Prosody, 010301 acoustics, business.industry, [INFO.INFO-MM]Computer Science [cs]/Multimedia [cs.MM], [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], 020206 networking & telecommunications, Human-Computer Interaction, Formant, [INFO.INFO-SD]Computer Science [cs]/Sound [cs.SD], Mimicry, automatic target speaker selection, crowdsourcing, business, Cryptography and Security (cs.CR), [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Software, mimicry, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

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

Published
2019
Full Text
View/download PDF

32. Can We Use Speaker Recognition Technology to Attack Itself? Enhancing Mimicry Attacks Using Automatic Target Speaker Selection

Author

Rosa González Hautamäki, Sahidullah, Tomi Kinnunen, Ville Vestman, University of Eastern Finland, Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), and Sahidullah, Md

Subjects
FOS: Computer and information sciences, [INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI], Spoofing attack, [INFO.INFO-TS] Computer Science [cs]/Signal and Image Processing, Computer science, Speech recognition, Transferability, Machine Learning (stat.ML), Context (language use), 02 engineering and technology, spoofing, 01 natural sciences, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], [INFO.INFO-CV] Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Audio and Speech Processing (eess.AS), Statistics - Machine Learning, 020204 information systems, 0103 physical sciences, FOS: Electrical engineering, electronic engineering, information engineering, 0202 electrical engineering, electronic engineering, information engineering, Selection (linguistics), [INFO.INFO-HC]Computer Science [cs]/Human-Computer Interaction [cs.HC], 010301 acoustics, [SPI.SIGNAL] Engineering Sciences [physics]/Signal and Image processing, [SPI.ACOU]Engineering Sciences [physics]/Acoustics [physics.class-ph], [INFO.INFO-MM] Computer Science [cs]/Multimedia [cs.MM], [SPI.ACOU] Engineering Sciences [physics]/Acoustics [physics.class-ph], Computer Science - Computation and Language, [INFO.INFO-MM]Computer Science [cs]/Multimedia [cs.MM], [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], [INFO.INFO-LG] Computer Science [cs]/Machine Learning [cs.LG], [SCCO.LING]Cognitive science/Linguistics, Speaker recognition, [PHYS.MECA.ACOU]Physics [physics]/Mechanics [physics]/Acoustics [physics.class-ph], Mimicry, Speaker verification, [INFO.INFO-HC] Computer Science [cs]/Human-Computer Interaction [cs.HC], [SCCO.LING] Cognitive science/Linguistics, [PHYS.MECA.ACOU] Physics [physics]/Mechanics [physics]/Acoustics [physics.class-ph], Computation and Language (cs.CL), [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, mimicry, Electrical Engineering and Systems Science - Audio and Speech Processing
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., (A slightly shorter version) has been submitted to IEEE ICASSP 2019

Published
2019

33. Who Do I Sound like? Showcasing Speaker Recognition Technology by Youtube Voice Search

Author

Anssi Kanervisto, Bilal Soomro, Tomi Kinnunen, Ville Hautamäki, and Ville Vestman

Subjects
FOS: Computer and information sciences, Sound (cs.SD), Voice search, business.industry, Computer science, Event (computing), 02 engineering and technology, Speaker recognition, computer.software_genre, Computer Science - Sound, World Wide Web, 030507 speech-language pathology & audiology, 03 medical and health sciences, Identification (information), Leverage (negotiation), Audio and Speech Processing (eess.AS), FOS: Electrical engineering, electronic engineering, information engineering, 0202 electrical engineering, electronic engineering, information engineering, Web application, 020201 artificial intelligence & image processing, Web service, 0305 other medical science, business, computer, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

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

Published
2019
Full Text
View/download PDF

34. Effects of Gender Information in Text-Independent and Text-Dependent Speaker Verification

Author

Ville Vestman, Sahidullah, Ville Hautamäki, Tomi Kinnunen, Anssi Kanervisto, and School of Computing, activities

Subjects
Speaker verification, ComputingMilieux_THECOMPUTINGPROFESSION, Computer science, Speech recognition, Text independent, 020206 networking & telecommunications, 02 engineering and technology, Speaker recognition, Speaker diarisation, gender dependent system, 030507 speech-language pathology & audiology, 03 medical and health sciences, gender classification, 0202 electrical engineering, electronic engineering, information engineering, NIST, Mel-frequency cepstrum, 0305 other medical science, Classifier (UML), speaker verification
Abstract

It is well-known that for speaker recognition task, gender-dependent acoustic modeling performs better than gender-independent modeling. The practice is to use the gender ground-truth and to train gender-dependent models. However, such information is not necessarily available, especially if speakers are remotely enrolled. A way to overcome this is to use a gender classification system, which introduces an additional layer of uncertainty. To date, such uncertainty has not been studied. We implement two gender classifier systems and test them with two different corpora and speaker verification systems. We find that estimated gender information can improve speaker verification accuracy over gender-independent methods. Our detailed analysis suggests that gender estimation should have a sufficiently high accuracy to yield improvements in speaker verification performance., final draft, peerReviewed

Published
2017

Catalog

Books, media, physical & digital resources
See catalog results