Multimodal voice phishing and synthetic voice detection on Synthetic audios
73AccuracyCNN-BiLSTM with MFCC features
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CNN-BiLSTM with MFCC featuresComplexity=medium–high complexity, Audio duration=20–170s synthetic audios, Mode=Voice-only2026.06 | 73 | 32.4 |