V

Voxlingua107 Epaca Tdnn

Developed by TalTechNLP
ECAPA-TDNN architecture spoken language identification model trained on the VoxLingua107 dataset, supporting recognition of 107 languages
Downloads 10.21k
Release Time : 3/2/2022

Model Overview

This model can identify the language used in speech segments and can also serve as a speech feature extractor for training specialized language identification models

Model Features

Multilingual Support
Supports recognition of 107 languages, covering major global languages
ECAPA-TDNN Architecture
Utilizes the advanced ECAPA-TDNN neural network architecture, previously successfully applied to speaker recognition tasks
Dual Purpose
Can be directly used for language identification or as a feature extractor for training specialized models

Model Capabilities

Speech Language Identification
Speech Feature Extraction
Multilingual Processing

Use Cases

Speech Technology
Language Identification Service
Provides language identification functionality for voice applications
7% error rate on the development dataset
Speech Feature Extraction
Extracts speech segment features for training specialized language identification models
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase