A

Asr Transformer Transformerlm Librispeech

Developed by speechbrain
This is an automatic speech recognition (ASR) system based on Transformer architecture, combining CTC and Transformer decoder, trained on the LibriSpeech English dataset.
Downloads 533
Release Time : 3/2/2022

Model Overview

This model is an end-to-end automatic speech recognition system that includes a tokenizer, neural language model, and acoustic model, supporting English speech transcription.

Model Features

Joint Decoding
Combines CTC probabilities with Transformer decoder for joint decoding, improving recognition accuracy
Subword Unit Processing
Uses unigram tokenizer to convert words into subword units, enhancing model processing capability
High Performance
Achieves word error rates (WER) of 2.27 (clean) and 5.53 (other) on LibriSpeech test sets

Model Capabilities

English Speech Recognition
Audio Transcription
Automatic Speech Recognition

Use Cases

Speech Transcription
Audio File Transcription
Convert English speech files to text
Highly accurate transcription results
Speech Processing
Speech Recognition System
Integration into speech recognition applications
Provides accurate speech-to-text functionality
Featured Recommended AI Models
ยฉ 2025AIbase