Audio Language Model for Deepfake Detection Grounded in Acoustic Chain-of-Thought

TL;DR

This paper presents CoLMbo-DF, an audio language model that enhances deepfake speech detection by integrating structured acoustic evidence and reasoning, leading to improved accuracy and interpretability.

Contribution

It introduces a novel feature-guided audio language model with explicit acoustic reasoning and a new dataset with chain-of-thought annotations for deepfake detection.

Findings

Outperforms existing baselines in deepfake detection accuracy.

Grounds model reasoning in interpretable acoustic evidence.

Uses a lightweight language model with significant improvements.

Abstract

Deepfake speech detection systems are often limited to binary classification tasks and struggle to generate interpretable reasoning or provide context-rich explanations for their decisions. These models primarily extract latent embeddings for authenticity detection but fail to leverage structured acoustic evidence such as prosodic, spectral, and physiological attributes in a meaningful manner. This paper introduces CoLMbo-DF, a Feature-Guided Audio Language Model that addresses these limitations by integrating robust deepfake detection with explicit acoustic chain-of-thought reasoning. By injecting structured textual representations of low-level acoustic features directly into the model prompt, our approach grounds the model's reasoning in interpretable evidence and improves detection accuracy. To support this framework, we introduce a novel dataset of audio pairs paired with chain-of-thought annotations. Experiments show that our method, trained on a lightweight open-source language model, significantly outperforms existing audio language model baselines despite its smaller scale, marking a significant advancement in explainable deepfake speech detection.