W

Wav2vec2 Large 960h Lv60 Self En Atc Uwb Atcc

Developed by Jzuluaga
A speech recognition model fine-tuned on the UWB-ATCC Air Traffic Control Communication dataset based on the facebook/wav2vec2-large-960h-lv60-self model
Downloads 115
Release Time : 11/30/2022

Model Overview

This model is specifically designed for English speech recognition in Air Traffic Control (ATC) communication scenarios, excelling in domain shift situations

Model Features

Strong Domain Adaptability
Optimized for Air Traffic Control communication scenarios, maintaining high performance even under domain shifts
Low Resource Efficiency
Significant performance improvement achieved with only a small amount of labeled data for fine-tuning
Supports Language Model Integration
Can be combined with KenLM language model to further enhance recognition accuracy

Model Capabilities

English Air Traffic Control Speech Recognition
Speech Recognition in Noisy Environments
Domain-Specific Terminology Recognition

Use Cases

Air Traffic Control
ATC Communication Transcription
Converting voice communications between air traffic controllers and pilots into text
Word Error Rate (WER) 17.2 (without LM)/13.72 (with LM)
Aviation Voice Data Analysis
Supporting aviation safety research and communication efficiency analysis
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase