Attentive Merging of Hidden Embeddings from Pre-trained Speech Model for Anti-spoofing Detection

TL;DR

It is found that the early hidden transformer layers of the WavLM large model contribute significantly to anti-spoofing task, enabling computational efficiency by utilizing a partial pre-trained model.

Abstract

Self-supervised learning (SSL) speech representation models, trained on large speech corpora, have demonstrated effectiveness in extracting hierarchical speech embeddings through multiple transformer layers. However, the behavior of these embeddings in specific tasks remains uncertain. This paper investigates the multi-layer behavior of the WavLM model in anti-spoofing and proposes an attentive merging method to leverage the hierarchical hidden embeddings. Results demonstrate the feasibility of fine-tuning WavLM to achieve the best equal error rate (EER) of 0.65%, 3.50%, and 3.19% on the ASVspoof 2019LA, 2021LA, and 2021DF evaluation sets, respectively. Notably, We find that the early hidden transformer layers of the WavLM large model contribute significantly to anti-spoofing task, enabling computational efficiency by utilizing a partial pre-trained model.

Figures (2)

• Figure 1: Diagram of the attentive merging method for the multi-layer hidden embeddings from WavLM pre-trained model (illustrated for 6 layers).
• Figure 2: Normalized linear weights on the hidden embeddings from the 24 transformer encoders in the WavLM large model.