ASVspoof 2019:

Automatic Speaker Verification

Spoofing and Countermeasures Challenge


Previous challenges: If you're searching for the ASVspoof 2017 website, it is still available here. The ASVspoof 2015 website is still available here

The evaluation plan for 2019 is under active development. It will be made available to all via the asvspoof.org website as soon as it is complete.

The tentative schedule for the 2019 evaluation is as follows:

late November to early December, 2018Development data
early-to-mid February, 2019Evaluation data
+1 weekScore submission
end of February, 2019 Results
March 29, 2019Interspeech deadline

Those interested in ASVspoof may wish to read our Odyssey 2018 paper which describes the t-DCF metric that we intend to adopt for the ASVspoof 2019 evaluation:

T. Kinnunen, K.-A. Lee, H. Delgado, N. Evans, M. Todisco, M. Sahidullah, J. Yamagishi, D.-A. Reynolds, ”t-DCF: a Detection Cost Function for the Tandem Assessment of Spoofing Countermeasures and Automatic Speaker Verification ”, Proc. Odyssey 2018 - The Speaker and Language Recognition Workshop [PDF]

A software implementation of the t-DCF metric can be downloaded here.

Organisers

Tomi Kinnunen, University of Eastern Finland, FINLAND
Nicholas Evans, EURECOM, FRANCE
Junichi Yamagishi, National Institute of Informatics, JAPAN / University of Edinburgh, UK
Kong Aik Lee, NEC, JAPAN
Md Sahidullah, Inria, FRANCE
Massimiliano Todisco, EURECOM, FRANCE
Héctor Delgado, EURECOM, FRANCE