A new hybrid spoofed-audio benchmark is claimed to show that fine-tuning on it reaches 97%+ accuracy, but the reported numbers are internally inconsistent.
Towards Attention-based Contrastive Learning for Audio Spoof Detection
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Vision transformers (ViT) have made substantial progress for classification tasks in computer vision. Recently, Gong et. al. '21, introduced attention-based modeling for several audio tasks. However, relatively unexplored is the use of a ViT for audio spoof detection task. We bridge this gap and introduce ViTs for this task. A vanilla baseline built on fine-tuning the SSAST (Gong et. al. '22) audio ViT model achieves sub-optimal equal error rates (EERs). To improve performance, we propose a novel attention-based contrastive learning framework (SSAST-CL) that uses cross-attention to aid the representation learning. Experiments show that our framework successfully disentangles the bonafide and spoof classes and helps learn better classifiers for the task. With appropriate data augmentations policy, a model trained on our framework achieves competitive performance on the ASVSpoof 2021 challenge. We provide comparisons and ablation studies to justify our claim.
citation-role summary
citation-polarity summary
fields
cs.SD 1years
2025 1verdicts
REJECT 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
When Fine-Tuning is Not Enough: Lessons from HSAD on Hybrid and Adversarial Audio Spoof Detection
A new hybrid spoofed-audio benchmark is claimed to show that fine-tuning on it reaches 97%+ accuracy, but the reported numbers are internally inconsistent.