Towards Explainable Spoofed Speech Attribution and Detection:a Probabilistic Approach for Characterizing Speech Synthesizer Components

TL;DR

This paper introduces an explainable probabilistic framework for spoofed speech detection and attribution, decomposing speech into interpretable attribute embeddings that perform comparably to raw embeddings on the ASVspoof2019 dataset.

Contribution

The paper presents a novel probabilistic attribute embedding approach that enhances interpretability while maintaining high performance in spoofing detection and source attribution.

Findings

Achieves 99.7% accuracy and 0.22% EER in spoofing detection.

Attains 90.23% accuracy and 2.07% EER in attack attribution.

Demonstrates interpretability through Shapley value analysis.

Abstract

We propose an explainable probabilistic framework for characterizing spoofed speech by decomposing it into probabilistic attribute embeddings. Unlike raw high-dimensional countermeasure embeddings, which lack interpretability, the proposed probabilistic attribute embeddings aim to detect specific speech synthesizer components, represented through high-level attributes and their corresponding values. We use these probabilistic embeddings with four classifier back-ends to address two downstream tasks: spoofing detection and spoofing attack attribution. The former is the well-known bonafide-spoof detection task, whereas the latter seeks to identify the source method (generator) of a spoofed utterance. We additionally use Shapley values, a widely used technique in machine learning, to quantify the relative contribution of each attribute value to the decision-making process in each task. Results on the ASVspoof2019 dataset demonstrate the substantial role of duration and conversion modeling in spoofing detection; and waveform generation and speaker modeling in spoofing attack attribution. In the detection task, the probabilistic attribute embeddings achieve $99.7\%$ balanced accuracy and $0.22\%$ equal error rate (EER), closely matching the performance of raw embeddings ($99.9\%$ balanced accuracy and $0.22\%$ EER). Similarly, in the attribution task, our embeddings achieve $90.23\%$ balanced accuracy and $2.07\%$ EER, compared to $90.16\%$ and $2.11\%$ with raw embeddings. These results demonstrate that the proposed framework is both inherently explainable by design and capable of achieving performance comparable to raw CM embeddings.