W

Wav2vec2 Large Xlsr 53 Dutch

Developed by jonatasgrosman
A Dutch speech recognition model fine-tuned based on facebook/wav2vec2-large-xlsr-53, trained on the Common Voice and CSS10 datasets, supporting 16kHz audio input.
Downloads 3.0M
Release Time : 3/2/2022

Model Overview

An automatic speech recognition (ASR) model optimized for Dutch, based on the XLSR-53 architecture, suitable for converting Dutch speech to text.

Model Features

Multi-dataset Training
Fine-tuned using the training and validation sets from Common Voice 6.1 and CSS10 datasets.
Language Model Support
Can be used with a language model to significantly improve recognition accuracy (WER reduced from 15.72 to 12.84).
Efficient Speech Processing
Supports 16kHz sampling rate audio input, suitable for real-time speech recognition scenarios.

Model Capabilities

Dutch speech recognition
Audio-to-text conversion
Enhanced recognition with language model support

Use Cases

Speech Transcription
Dutch Speech Transcription
Convert Dutch speech content into text format
Test set WER 15.72/CER 5.35
Voice Assistants
Dutch Voice Command Recognition
Used for Dutch voice command recognition in voice assistants or smart home devices
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase