Year: 2026
Type: conference-paper
Abstract: Advances in speech synthesis have made deepfake speeches increasingly convincing, posing growing threats to security.While self-supervised learning (SSL) based detectors achieve state-of-the-art performance, their computational demands (typically 300M+ parameters) prevent deployment on resourceconstrained devices.Existing compression methods, designed mainly for content-centric tasks, struggle to maintain competitive performance when directly adapted to deepfake detection.We propose a Task-Aware Joint Pruning and Distillation framework that combines cross-domain knowledge distillation with movement-guided structured pruning to transfer forgerydiscriminative knowledge and preserve critical structures under aggressive compression.Our framework reduces the model to 31.9M parameters with 6.3× FLOPs reduction, with an average performance drop of only 1.30% across multiple datasets compared to the uncompressed baseline, demonstrating strong potential for on-device deployment.
Source: Interspeech (USB)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7220724685