AST-VoxCelebSpoof Synthetic Speech Detection Model - Open-source and Free for Accurate Identification of Synthetic Speech

AST VoxCelebSpoof Synthetic Voice Detection

Developed by MattyB95

A synthetic speech detection model fine-tuned based on MIT/ast-finetuned-audioset-10-10-0.4593, demonstrating outstanding performance on the VoxCelebSpoof dataset

Audio Classification

Transformers

EnglishOpen Source License:MIT #High-precision voice detection #Synthetic speech recognition #Voiceprint anti-counterfeiting

Downloads 9,518

Release Time : 1/16/2024

Model Overview

This model is used to detect synthetic speech, fine-tuned on audio classification tasks based on the AST architecture, specifically optimized for voice spoofing detection scenarios

Model Features

High accuracy

Achieves 99.99% accuracy and F1 score on the evaluation set

AST-based architecture

Utilizes the Audio Spectrogram Transformer architecture, excelling in processing audio spectral features

Specialized for synthetic speech detection

Optimized for the VoxCelebSpoof dataset, particularly suitable for voice spoofing detection scenarios

Model Capabilities

Audio classification

Synthetic speech detection

Voice spoofing recognition

Use Cases

Security verification

Voice authentication systems

Used to detect synthetic speech attacks in voice authentication systems

Can effectively identify 99.99% of synthetic speech samples

Content moderation

Fake audio detection

Identifies AI-generated fake audio content on social media

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
2218896740319.232	1.0	29527	611463921664.0	0.9998	0.9998	0.9999	0.9997
522149441830.912	2.0	59054	284563668992.0	0.9997	0.9997	0.9999	0.9996
0.0	3.0	88581	89136693248.0	0.9999	0.9999	1.0	0.9998

Property	Details
Model Type	Fine - tuned version of [MIT/ast - finetuned - audioset - 10 - 10 - 0.4593](https://huggingface.co/MIT/ast - finetuned - audioset - 10 - 10 - 0.4593)
Training Data	MattyB95/VoxCelebSpoof
Metrics	accuracy, f1, precision, recall

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

AST VoxCelebSpoof Synthetic Voice Detection

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 AST-VoxCelebSpoof-Synthetic-Voice-Detection

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

🔧 Technical Details

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License