AI RESEARCH

Audio Deepfake Detection with Half-Truth Localisation Using Cross-Attentive Feature Fusion

arXiv CS.LG

ArXi:2605.29531v1 Announce Type: cross Audio deepfake detection is well-studied as a binary problem, but partially manipulated speech, where a short synthesised segment is spliced into an otherwise genuine utterance, poses a harder and realistic threat. Detecting such half-truth audio requires not only distinguishing it from real and fully fake speech, but also localising where the manipulation occurs.