A Data-Centric Approach to Generalizable Speech Deepfake Detection

TL;DR

This paper emphasizes the importance of data composition in speech deepfake detection, introducing a data-centric framework with empirical analysis and a novel sampling strategy that enhances model robustness and efficiency.

Contribution

It presents a comprehensive data-centric approach, including data scaling laws and the DOSS framework, to improve generalization in speech deepfake detection.

Findings

DOSS-Select outperforms naive data aggregation with only 3% data usage.

Training on 12k-hour data with DOSS-Weight achieves state-of-the-art results.

Data composition significantly impacts deepfake detection performance.

Abstract

Achieving robust generalization in speech deepfake detection (SDD) remains a primary challenge, as models often fail to detect unseen forgery methods. While research has focused on model-centric and algorithm-centric solutions, the impact of data composition is often underexplored. This paper proposes a data-centric approach, analyzing the SDD data landscape from two practical perspectives: constructing a single dataset and aggregating multiple datasets. To address the first perspective, we conduct a large-scale empirical study to characterize the data scaling laws for SDD, quantifying the impact of source and generator diversity. To address the second, we propose the Diversity-Optimized Sampling Strategy (DOSS), a principled framework for mixing heterogeneous data with two implementations: DOSS-Select (pruning) and DOSS-Weight (re-weighting). Our experiments show that DOSS-Select outperforms the naive aggregation baseline while using only 3% of the total available data. Furthermore, our final model, trained on a 12k-hour curated data pool using the optimal DOSS-Weight strategy, achieves state-of-the-art performance, outperforming large-scale baselines with greater data and model efficiency on both public benchmarks and a new challenge set of various commercial APIs.