G

Gender Cls Svm Ecapa Voxceleb

Developed by griko
Based on SpeechBrain's ECAPA-TDNN speaker embedding model and SVM classifier, it can predict speaker gender from audio input.
Downloads 29
Release Time : 11/9/2024

Model Overview

This model combines ECAPA-TDNN speaker embeddings with an SVM classifier to identify speaker gender from audio, supporting binary classification (male/female).

Model Features

High-precision classification
Achieves 98.9% accuracy on the VoxCeleb2 test set and 99.6% accuracy on the TIMIT test set.
Multi-dataset validation
Performance validated on VoxCeleb2, Mozilla Common Voice, and TIMIT datasets.
Optimized classifier
SVM classifier fine-tuned through 200 Optuna optimizations.
Automatic preprocessing
Supports automatic audio format conversion (16kHz/mono) and voice activity detection.

Model Capabilities

Gender classification
Speaker feature extraction
Audio processing
Voiceprint analysis

Use Cases

Speech analysis
Speaker gender recognition
Automatically identifies speaker gender from audio.
High accuracy (VoxCeleb2: 98.9%)
Speech dataset processing
Dataset gender labeling
Automatically adds gender labels to unlabeled speech datasets.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase