D

Deepfake Audio Detection V1

Developed by Zeyadd-Mostaffa
A deepfake audio detection model fine-tuned based on wav2vec2-base, achieving 99.66% accuracy
Downloads 33
Release Time : 2/10/2025

Model Overview

This model is specifically designed for detecting deepfake audio, fine-tuned from Facebook's wav2vec2-base model, and performs excellently on evaluation datasets.

Model Features

High Accuracy
Achieves 99.66% accuracy on evaluation datasets
Fine-tuned on wav2vec2
Leverages the powerful wav2vec2-base model for fine-tuning, inheriting its excellent audio feature extraction capabilities
Efficient Training
Requires only 5 training epochs to achieve high performance

Model Capabilities

Audio Classification
Deepfake Detection
Audio Feature Extraction

Use Cases

Security Detection
Scam Call Identification
Detects whether deepfake voices are used in phone calls
Effectively identifies over 99% of fake voices
Media Content Moderation
Detects whether audio content has been processed with deepfake technology
High accuracy in identifying fake audio
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase