W

Wav2vec2 Live Japanese

Developed by ttop324
A Japanese speech recognition model fine-tuned based on facebook/wav2vec2-large-xlsr-53, supporting hiragana output
Downloads 20
Release Time : 3/2/2022

Model Overview

This is an optimized Automatic Speech Recognition (ASR) model for Japanese, capable of converting Japanese speech into hiragana text. The model has been fine-tuned on multiple Japanese speech datasets and is suitable for Japanese speech transcription tasks.

Model Features

Multi-dataset Fine-tuning
Fine-tuned on multiple Japanese speech datasets including common_voice, JSUT, CSS10, TEDxJP-10K, JVS, and JSSS
Hiragana Output
Specifically optimized for Japanese hiragana conversion, capable of outputting standardized hiragana text
High Performance
Achieved 21.48% WER and 9.82% CER on the Common Voice Japanese test set

Model Capabilities

Japanese Speech Recognition
Audio to Text
Hiragana Conversion

Use Cases

Speech Transcription
Japanese Speech to Text
Convert Japanese speech content into hiragana text
21.48% WER accuracy
Assistive Tools
Real-time Caption Generation
Generate real-time captions for Japanese videos or live streams
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase