Introduction

Facebook Automatic Speech Recognition provides great models and workflows built by Facebook (now known as Meta) that you can use in your apps to carry out automatic speech recognition (ASR).

You can use this model to easily and quickly convert German audio to German text (speech-to-text transcription). Simply upload a German audio file from your local computer or add a publicly accessible audio URL, and the model will output the transcribed text.

Facebook ASR models will help you to effectively transcribe audio content into written words without having to type them manually. These models are also valuable to persons with disabilities who cannot use a keyboard.

Wav2Vec2-Large-XLSR-53

The base model pretrained and fine-tuned on 960 hours of Librispeech on 16kHz sampled speech audio.

When using the model make sure that your speech input is also sampled at 16Khz.

More Info

• Meta AI Research post: Wav2vec 2.0: Learning the structure of speech from raw audio
• Hugging Face docs: facebook/wav2vec2-large-xlsr-53-german

Paper

Unsupervised Cross-Lingual Representation Learning for Speech Recognition

Authors: Alexis Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli

Abstract

This paper presents XLSR which learns cross-lingual speech representations by pretraining a single model from the raw waveform of speech in multiple languages. We build on wav2vec 2.0 which is trained by solving a contrastive task over masked latent speech representations and jointly learns a quantization of the latents shared across languages. The resulting model is fine-tuned on labeled data and experiments show that cross-lingual pretraining significantly outperforms monolingual pretraining. On the CommonVoice benchmark, XLSR shows a relative phoneme error rate reduction of 72% compared to the best known results. On BABEL, our approach improves word error rate by 16% relative compared to a comparable system. Our approach enables a single multilingual speech recognition model which is competitive to strong individual models. Analysis shows that the latent discrete speech representations are shared across languages with increased sharing for related languages. We hope to catalyze research in low-resource speech understanding by releasing XLSR-53, a large model pretrained in 53 languages.

The XLSR Approach

A shared quantization module over feature encoder representations produces multilingual quantized speech units whose embeddings are then used as targets for a Transformer trained by contrastive learning. The model learns to share discrete tokens across languages, creating bridges across languages.

Test – Evaluation on Common Voice German

Result (WER): 18.5 %