S

Stt Ca Es Conformer Transducer Large

Developed by projecte-aina
A Catalan-Spanish bilingual automatic speech recognition model based on the NVIDIA Spanish model
Downloads 1,127
Release Time : 11/20/2024

Model Overview

This model is a bilingual automatic speech recognition (ASR) solution suitable for Catalan and Spanish. It is built on the Conformer-Transducer architecture of NVIDIA and can transcribe speech into pure text without punctuation.

Model Features

Bilingual support
Capable of handling speech recognition tasks in both Catalan and Spanish simultaneously
Large-scale training
Fine-tuned on a bilingual dataset totaling 7426 hours
High-performance architecture
Adopts the large variant architecture of Conformer-Transducer, with powerful speech recognition capabilities

Model Capabilities

Catalan speech recognition
Spanish speech recognition
Speech-to-text

Use Cases

Speech transcription
Meeting minutes
Transcribe Catalan or Spanish meeting recordings into text
Generate a punctuation-free pure text transcription result
Media content processing
Process speech in media content such as broadcasts and podcasts
Generate a written record for the media content
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase