AIbase
EN
Home
Wav2vec2 Xls R 300m Mixed
W

Wav2vec2 Xls R 300m Mixed

Developed by mesolitica
A speech recognition model fine-tuned on mixed-language datasets based on Facebook's wav2vec2-xls-r-300m model, supporting Malay, Singaporean English, and Mandarin.
Speech Recognition
Transformers
#Multilingual speech recognition#Mixed-language fine-tuning#Low CER/WER
Downloads 10.07k
Release Time : 6/1/2022

Model Overview

This model is a speech recognition model fine-tuned for three languages (Malay, Singaporean English, and Mandarin), suitable for multilingual speech-to-text tasks.

Model Features

Multilingual support
Supports speech recognition in three languages: Malay, Singaporean English, and Mandarin.
High performance
Performs excellently on evaluation sets with low Character Error Rate (CER) and Word Error Rate (WER).
Language model enhancement
Supports integration with external language models to further improve recognition accuracy.

Model Capabilities

Speech recognition
Multilingual processing
Speech-to-text

Use Cases

Speech transcription
Multilingual meeting minutes
Used to transcribe meeting content involving Malay, Singaporean English, and Mandarin.
Accurate transcription of mixed-language meeting content
Customer service dialogue analysis
Analyze multilingual customer service conversations.
Improves efficiency in customer service quality analysis
Education
Language learning assistance
Helps learners practice and evaluate pronunciation accuracy.
Provides instant pronunciation feedback
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase