AITF Audio Deepfake Detection AST — Fine-tuned Model
Model fine-tuned dari MIT/ast-finetuned-audioset-10-10-0.4593 untuk deteksi audio deepfake.
Label
| Label | Keterangan |
|---|---|
| Real | Audio asli yang dibuat oleh manusia |
| Fake | Audio sintetis atau hasil generative AI |
Training Info
Base model : MIT/ast-finetuned-audioset-10-10-0.4593
Architecture : Audio Spectrogram Transformer (AST)
Task : Binary Audio Deepfake Detection
Framework : PyTorch + Hugging Face Transformers
Input : Audio mono 16 kHz
Output : Single-logit binary classification
Loss Function : BCEWithLogitsLoss
Classifier Head : Linear(768 → 1)
- Downloads last month
- -
Model tree for aitf-komdigi/KomdigiITS-86M-DFK-DeepfakeAudioClassification
Base model
MIT/ast-finetuned-audioset-10-10-0.4593