STC Antispoofing Systems for the ASVspoof2019 Challenge
By G. Lavrentyeva, Sergey Novoselov, Andzhukaev Tseren, Marina Volkova, Artem Gorlanov, Alexander Kozlov. Interspeech, 2019
305 citations recorded by the source corpus.
This paper describes the Speech Technology Center (STC) antispoofing systems submitted to the ASVspoof 2019 challenge. The ASVspoof2019 is the extended version of the previous challenges and includes 2 evaluation conditions: logical access use-case scenario with speech synthesis and voice conversion attack types and physical access use-case scenario with replay attacks. During the challenge we developed anti-spoofing solutions for both scenarios. The proposed systems are implemented using deep learning approach and are based on different types of acoustic features. We enhanced Light CNN architecture previously considered by the authors for replay attacks detection and which performed high spoofing detection quality during the ASVspoof2017 challenge. In particular here we investigate the efficiency of angular margin based softmax activation for training robust deep Light CNN classifier to solve the mentioned-above tasks. Submitted systems achieved EER of 1.86% in logical access scenario and 0.54% in physical access scenario on the evaluation part of the Challenge corpora. High performance obtained for the unknown types of spoofing attacks demonstrates the stability of the offered approach in both evaluation conditions.
View Paper: https://www.semanticscholar.org/paper/12c3c3ef97ed245268a98a7eb273314c59c3c054
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Xin Eric Wang
- Junichi Yamagishi
- Massimiliano Todisco
- Xuechen Liu
- Héctor Delgado
- N. Evans
- Tomi H. Kinnunen
- Jose Patino
Citations in the corpus
- End-to-End anti-spoofing with RawNet2 (2020, unknown): “T45 [27] 0.”
- Does Audio Deepfake Detection Generalize? (2022, unknown): “Another common architecture for audio spoof detection are LCNN-based learning models such as LCNN , LCNN-Attention , and LCNN-LSTM [12, 13, 14].”
- ASVspoof 2021: accelerating progress in spoofed and deepfake speech detection (2021, methodology): “The back end is based on the LCNN reported in [22], but incorporates LSTM layers and average pooling.”