Fahad A. Aloufi

Evaluation of Speaker Verification Models in Arabic Datasets

This study presents a systematic benchmark of six state-of-the-art, pre-trained Speaker-Verification (SV) models, ECAPA-TDNN, ECAPA2, WavLM-Large-SV, Pyannote-Embedding, TitaNet, and WeSpeaker, on Arabic speech, a language for which modern SV models remain largely unevaluated. All models were trained on predominantly E...