Replay Attack Detection Using Generalized Cross-Correlation of Stereo Signal

Yaguchi, Ryoya; Shiota, Sayaka; Kiya, Hitoshi

doi:10.23919/eusipco.2019.8902985

Cited by 7 publications

(7 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From all non-speech segments of the input utterance, the GCC scores Φ min and Φ avg are calculated by Eqs. (12), (13). By using all frames of speech segments in an input utterance, the LLR and CS were calculated under the manner of Sections 3.2 and 3.3, respectively.…”

Section: Methodsmentioning

confidence: 99%

“…It has been reported that GCC-based methods achieved high performances, especially in quiet situations [13]. However, this performance is situation-dependent.…”

Section: Score Fusion System 41 Motivationmentioning

confidence: 98%

“…The GCC-based method [13] focuses on the trajectories of the maximum GCC (max-GCC) in non-speech segments for spoofing detection. The max-GCC for each frame is defined as…”

Section: Spoofing Detection Using Maximum Gcc In Nonspeech Segmentsmentioning

confidence: 99%

“…Since the ASVspoof database was recorded by using single-channel microphones, almost all proposed countermeasures assume a single-channel situation. Meanwhile, since recording with multi-channel microphones has become easy, replay attack detection (RAD) systems assuming multi-channel recording have also been proposed [11], [12], [13]. In Ref.…”

Section: Introductionmentioning

confidence: 99%

“…In Ref. [13], we use generalized cross-correlation (GCC) [14] of stereo signals for RAD. GCC-based systems focus on non-speech segments, in which no sound is emitted from humans, but loudspeakers tend to generate some noise and non-perceptual signals, and these signals can be easily captured in non-speech segments.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Replay Attack Detection Based on Spatial and Spectral Features of Stereo Signal

Yaguchi

Shiota

Kiya

2021

Journal of Information Processing

Self Cite

View full text Add to dashboard Cite

In this paper, we propose a replay attack detection (RAD) method that uses spatial and spectral features of a stereo signal. To distinguish genuine and replayed utterance, we focus on non-speech segments, in which a human does not emit sound, but a loudspeaker for replay attack might emit some recorded noise or its electromagnetic noise. The generalized cross-correlation (GCC) based spatial features capture this difference. To improve the robustness against the variety of recording environments, we combine the spatial features with spectral features. In particular, we fuse the output scores of GCC-based and spectral feature-based methods. In experiments, we confirm the effectiveness of the combination of spatial and spectral features.

show abstract

Section: Methodsmentioning

confidence: 99%

“…It has been reported that GCC-based methods achieved high performances, especially in quiet situations [13]. However, this performance is situation-dependent.…”

Section: Score Fusion System 41 Motivationmentioning

confidence: 98%

“…The GCC-based method [13] focuses on the trajectories of the maximum GCC (max-GCC) in non-speech segments for spoofing detection. The max-GCC for each frame is defined as…”

Section: Spoofing Detection Using Maximum Gcc In Nonspeech Segmentsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations