A classifier's embedding space plus maximum cosine similarity to reference tracks verifies whether a speech deepfake came from the same generator, even for generators unseen during training.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Source Verification for Speech Deepfakes
A classifier's embedding space plus maximum cosine similarity to reference tracks verifies whether a speech deepfake came from the same generator, even for generators unseen during training.