ASVspoof 2015: the first automatic speaker verification spoofing and countermeasures challenge

By Zhizheng Wu, Tomi H. Kinnunen, N. Evans, Junichi Yamagishi, Cemal Hanilçi, Md. Sahidullah, Aleksandr Sizov. Interspeech, 2015

594 citations recorded by the source corpus.

An increasing number of independent studies have confirmed the vulnerability of automatic speaker verification (ASV) technology to spoofing. However, in comparison to that involving other biometric modalities, spoofing and countermeasure research for ASV is still in its infancy. A current barrier to progress is the lack of standards which impedes the comparison of results generated by different researchers. The ASVspoof initiative aims to overcome this bottleneck through the provision of standard corpora, protocols and metrics to support a common evaluation. This paper introduces the first edition, summaries the results and discusses directions for future challenges and research.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/8b4ee8c42155194c3e3f23eeb1f8f0f18f21731d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

A New Feature for Automatic Speaker Verification Anti-Spoofing: Constant Q Cepstral Coefficients (2016, result): “This view is supported by the results of the recent ASVspoof 2015 challenge [11] of which the winning system [12] utilised non-conventional features in conjunction with a classical Gaussian mixture model (GMM) classifier.”
STC Antispoofing Systems for the ASVspoof2019 Challenge (2019, result): “This behaviour differs from the earlier experiments on ASVspoof2015 [11] and ASVspoof2017 [12] datasets, that did not contain artificially produced data.”
One-Class Learning Towards Synthetic Voice Spoofing Detection (2020, background): “[6], [7] has been providing datasets and metrics for anti-spoofing speaker verification research.”
A comparison of features for synthetic speech detection (2015, methodology): “The overall best recognition accuracy on development set (EER of 0 .”
Constant Q cepstral coefficients: A spoofing countermeasure for automatic speaker verification (2017, methodology): “The most potentially damaging spooﬁng attacks in this case are voice conversion and speech synthesis (Wu et al., 2015).”
An Overview of Voice Conversion and Its Challenges: From Statistical Modeling to Deep Learning (2020, methodology): “This database became a reference standard in the automatic speaker verification (ASV) community [298], [299].”
The ASVspoof 2017 Challenge: Assessing the Limits of Replay Spoofing Attack Detection (2017, background): “ASVspoof 2017 follows two special sessions on spoofing and countermeasures for automatic speaker verification at INTERSPEECH 2013 [13] and 2015 [14] which formed the first evaluation, ASVspoof 2015 [15].”
ASVspoof 2019: Automatic Speaker Veriﬁcation Spooﬁng and Countermeasures Challenge Evaluation Plan ∗ (2021, background): “The ASVspoof 2019 challenge follows on from three special sessions on spoofing and countermeasures for automatic speaker verification held during INTERSPEECH 2013 [1], 2015 [2] and 2017 [3].”
rVAD: An Unsupervised Segment-Based Robust Voice Activity Detection Method (2019, methodology): “Brieﬂy, we compared the output of the SFT detector of diﬀerent values of θ sft with the output of the pitch detector on large number of utterances from various databases including NIST 2016 SRE (evaluation set) [60], TIMIT [61], RSR2015 [62], RedDots [38], ASVspoof2015 [63] and noisy versions (car, ”
ASVspoof 2019: Future Horizons in Spoofed and Fake Audio Detection (2019, background): “The ASVspoof initiative [1, 2, 3] spearheads research in antispoofing for automatic speaker verification (ASV).”
ASVspoof 2019: A large-scale public database of synthesized, converted and replayed speech (2019, background): “The first ASVspoof challenge held in 2015 focused on the development of countermeasures for the detection of speech synthesis and voice conversion spoofing attacks [5].”
ASVspoof 2021: Towards Spoofed and Deepfake Speech Detection in the Wild (2022, background): “The first edition of ASVspoof in 2015 [8] focused on the development of CMs for the detection of VC and TTS attacks.”
ASVspoof 2021: accelerating progress in spoofed and deepfake speech detection (2021, methodology): “ASVspoof 2021 is the forth edition in a series of biannual challenges [3, 4, 5] in which participants are tasked with the design of CMs for one or more tasks.”