X

Xls R Uzbek Cv8

Developed by lucio
An automatic speech recognition model fine-tuned on the Uzbek Common Voice 8 dataset based on facebook/wav2vec2-xls-r-300m
Downloads 2,860
Release Time : 3/2/2022

Model Overview

This model is an automatic speech recognition system for the Uzbek language, trained on the Common Voice 8 dataset, suitable for low-accuracy speech-to-text scenarios

Model Features

Uzbek-specific model
Speech recognition model specifically optimized for the Uzbek language
Low-resource optimization
Achieves good results using only 50% of the Common Voice dataset training data
Language model enhancement
Significantly improves recognition accuracy with kenlm language model integration
Special character handling
Correctly processes special characters <‘> and <’> in Uzbek

Model Capabilities

Uzbek speech recognition
Low-accuracy speech-to-text
Broadcast recording indexing

Use Cases

Media processing
Video subtitle generation
Generates preliminary subtitle drafts for Uzbek videos
Broadcast recording indexing
Automatically indexes Uzbek broadcast content
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase