ASVspoof 5: Design, collection and validation of resources for spoofing, deepfake, and adversarial attack detection using crowdsourced speech

ASVspoof 5 is the fifth edition in a series of challenges which promote the study of speech spoofing and deepfake attacks as well as the design of detection solutions. We introduce the ASVspoof 5 database which is generated in a crowdsourced fashion from data collected in diverse acoustic conditions (cf. studio-quality data for earlier ASVspoof databases) and from ∼ 2,000 speakers (cf. ∼ 100 earlier). The database contains attacks generated with 32 different algorithms, also crowdsourced, and optimised to varying degrees using new surrogate detection models. Among them are attacks generated with a mix of legacy and contemporary text-to-speech synthesis and voice conversion models, in addition to adversarial attacks which are incorporated for the first time. ASVspoof 5 protocols comprise seven speaker-disjoint partitions. They include two distinct partitions for the training of different sets of attack models, two more for the development and evaluation of surrogate detection models, and then three additional partitions which comprise the ASVspoof 5 training, development and evaluation sets. An auxiliary set of data collected from an additional 30k speakers can also be used to train speaker encoders for the implementation of attack algorithms. Also described herein is an experimental validation of the new ASVspoof 5 database using a set of automatic speaker verification and spoof/deepfake baseline detectors. With the exception of protocols and tools for the generation of spoofed/deepfake speech, the resources described in this paper, already used by participants of the ASVspoof 5 challenge in 2024, are now all freely available to the community. • Most advanced ASVspoof database for speech anti-spoofing and deepfake detection. • Speech data from more than 1,000 speakers in diverse acoustic conditions. • Speech deepfake data from 32 legacy and advanced synthesis and adversarial attacks. • Diverse encoding schemes and post-processing to address shortcut learnings. • Visualisation and experimental validation on the database.

Authors

Institutions

Publication Details

Journal
Computer Speech & Language
Published
2025-05-28
DOI
https://doi.org/10.1016/j.csl.2025.101825
Citations
8
Primary Topic
Speech Recognition and Synthesis
Type
article
Field-Weighted Citation Impact
19.85

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

ASVspoof 5: Design, collection and validation of resources for spoofing, deepfake, and adversarial attack detection using crowdsourced speech

Vishwanath P. Singh, Kong Aik Lee, Junichi Yamagishi, Soumi Maiti et al.
8 citations
Computer Speech & Language
Speech Recognition and Synthesis
19.85
article

ASVspoof 5: Design, collection and validation of resources for spoofing, deepfake, and adversarial attack detection using crowdsourced speech

Vishwanath P. Singh, Kong Aik Lee, Junichi Yamagishi, Soumi Maiti, Xuechen Liu, Hemlata Tak, Siwei Lyu, Sébastien Le Maguer, Tomi Kinnunen, Cheng Gong, Massimiliano Todisco, Chengxu Sun, Shuwei Hou, Nicholas Evans, Wangyou Zhang, Nicolas Müller, Héctor Delgado, Florian Lux, Md Sahidullah, Ivan Kukanov, Jee-weon Jung, Xin Wang, Liping Chen, Hanjie Guo, Ge Zhu, You Zhang, Myeonghun Jeong, Hye-jin Shim, Yongyi Zang
article en
8 citations

Abstract

ASVspoof 5 is the fifth edition in a series of challenges which promote the study of speech spoofing and deepfake attacks as well as the design of detection solutions. We introduce the ASVspoof 5 database which is generated in a crowdsourced fashion from data collected in diverse acoustic conditions (cf. studio-quality data for earlier ASVspoof databases) and from ∼ 2,000 speakers (cf. ∼ 100 earlier). The database contains attacks generated with 32 different algorithms, also crowdsourced, and optimised to varying degrees using new surrogate detection models. Among them are attacks generated with a mix of legacy and contemporary text-to-speech synthesis and voice conversion models, in addition to adversarial attacks which are incorporated for the first time. ASVspoof 5 protocols comprise seven speaker-disjoint partitions. They include two distinct partitions for the training of different sets of attack models, two more for the development and evaluation of surrogate detection models, and then three additional partitions which comprise the ASVspoof 5 training, development and evaluation sets. An auxiliary set of data collected from an additional 30k speakers can also be used to train speaker encoders for the implementation of attack algorithms. Also described herein is an experimental validation of the new ASVspoof 5 database using a set of automatic speaker verification and spoof/deepfake baseline detectors. With the exception of protocols and tools for the generation of spoofed/deepfake speech, the resources described in this paper, already used by participants of the ASVspoof 5 challenge in 2024, are now all freely available to the community. • Most advanced ASVspoof database for speech anti-spoofing and deepfake detection. • Speech data from more than 1,000 speakers in diverse acoustic conditions. • Speech deepfake data from 32 legacy and advanced synthesis and adversarial attacks. • Diverse encoding schemes and post-processing to address shortcut learnings. • Visualisation and experimental validation on the database.

Computer Speech & LanguageVol. 95
University of Stuttgart (DE), Microsoft (United States) (US), Seoul National University (KR), Hong Kong Polytechnic University (HK), National Institute of Informatics (JP), EURECOM (FR), Northrop Grumman (United States) (US), Joensuu Science Park (FI), Erlang Solutions (United Kingdom) (GB), Gidropribor (RU), Fraunhofer Institute for Applied and Integrated Security (DE), TCG Crest (IN), University of Rochester (US), Carnegie Mellon University (US)
Agence Nationale de la Recherche, Academia Sinica, Renmin University of China, China Scholarship Council, National Taiwan University, Precursory Research for Embryonic Science and Technology
Openalex Percentile: Top 1%
Speech Recognition and Synthesis
19.85
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.