works on

From the 1 of 10 linked papers with an AI index.

activity
20242026
collaborators
Showing cs.SDShow all

6 papers · 1 filter

cs.SD2026

Evidence Subspace Projection: Measuring How Much Evidence Explains Deepfake Detection in Self-Supervised Speech Models

Yixuan Xiao, Cheng-Wei Lin, Xin Wang +5

The paper introduces Evidence Subspace Projection, a technique that quantifies how different evidence factors (like attack type, codec, gender, transmission) explain the decisions…

cs.SD2026

DeepFense: A Unified, Modular, and Extensible Framework for Robust Deepfake Audio Detection

Yassine El Kheir, Arnab Das, Yixuan Xiao +6

Speech deepfake detection is a well-established research field with different models, datasets, and training strategies. However, the lack of standardized implementations and evalu…

cs.SD2026

DFKI-Speech System for WildSpoof Challenge: A robust framework for SASV In-the-Wild

Arnab Das, Yassine El Kheir, Enes Erdem Erdogan +3

This paper presents the DFKI-Speech system developed for the WildSpoof Challenge under the Spoofing aware Automatic Speaker Verification (SASV) track. We propose a robust SASV fram…

cs.SD2025

A Parameter-Efficient Multi-Scale Convolutional Adapter for Synthetic Speech Detection

Yassine El Kheir, Fabian Ritter-Guttierez, Arnab Das +2

Recent synthetic speech detection models typically adapt a pre-trained SSL model via finetuning, which is computationally demanding. Parameter-Efficient Fine-Tuning (PEFT) offers a…

cs.SD2025

Generalizable Audio Spoofing Detection using Non-Semantic Representations

Arnab Das, Yassine El Kheir, Carlos Franzreb +3

Rapid advancements in generative modeling have made synthetic audio generation easy, making speech-based services vulnerable to spoofing attacks. Consequently, there is a dire need…

cs.SD2025

Two Views, One Truth: Spectral and Self-Supervised Features Fusion for Robust Speech Deepfake Detection

Yassine El Kheir, Arnab Das, Enes Erdem Erdogan +3

Recent advances in synthetic speech have made audio deepfakes increasingly realistic, posing significant security risks. Existing detection methods that rely on a single modality,…