W

Wav2vec2 Xls R 1b Dutch

Developed by jonatasgrosman
This is a Dutch automatic speech recognition (ASR) model fine-tuned based on the XLS-R 1 billion parameter model, trained on multiple datasets including Common Voice 8.0, supporting 16kHz sampling rate audio input.
Downloads 146
Release Time : 3/2/2022

Model Overview

This model is specifically designed for Dutch automatic speech recognition tasks, fine-tuned from Facebook's XLS-R 1B parameter model, and performs excellently on multiple Dutch speech datasets.

Model Features

High-performance Dutch recognition
Achieves 10.38% WER and 3.04% CER on the Common Voice 8.0 test set
Supports language model enhancement
When combined with a language model, WER can be reduced to 6.83% and CER to 2.31%
Multi-dataset training
Trained on multiple datasets including Common Voice 8.0, Multilingual LibriSpeech, and Voxpopuli

Model Capabilities

Dutch speech recognition
16kHz audio processing
High-accuracy transcription

Use Cases

Speech transcription
Dutch speech-to-text
Convert Dutch speech content into text
Achieves over 90% accuracy on standard test sets
Voice assistants
Dutch voice command recognition
Used for Dutch voice assistant or smart home device command recognition
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase