ASVspoof-2021

ASVspoof2021 is a benchmark dataset used to evaluate the performance of automatic speaker verification systems against sophisticated spoofing attacks, particularly focusing on synthetic speech detection.

Papers using ASVspoof-2021 (13)

The Vicomtech Audio Deepfake Detection System Based On Wav2vec2 For The 2022 ADD Challenge2022 · 78 citesTemporal-channel Modeling In Multi-head Self-attention For Synthetic Speech Detection2024 · 31 citesCharacterizing The Temporal Dynamics Of Universal Speech Representations For Generalizable Deepfake Detection2023 · 7 citesXLSR-Kanformer: A KAN-Intergrated model for Synthetic Speech Detection2025 · 1 citesPushing the Performance of Synthetic Speech Detection with Kolmogorov-Arnold Networks and Self-Supervised Learning Models2025 · 2 citesAmplifying Artifacts with Speech Enhancement in Voice Anti-spoofing2025Collaborative Watermarking For Adversarial Speech Synthesis2023Mixture Of Experts Fusion For Fake Audio Detection Using Frozen Wav2vec 2.02024Improved DeepFake Detection Using Whisper Features2023 · 2 citesCharacterizing the temporal dynamics of universal speech representations for generalizable deepfake detection2023 · 1 citesCollaborative Watermarking for Adversarial Speech Synthesis2023Temporal-Channel Modeling in Multi-head Self-Attention for Synthetic Speech Detection2024Mixture of Experts Fusion for Fake Audio Detection Using Frozen wav2vec 2.02024