File size: 1,042 Bytes
346b014 9139449 346b014 9139449 346b014 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 |
---
license: apache-2.0
language:
- en
pipeline_tag: automatic-speech-recognition
datasets:
- LRS3
tags:
- Audio Visual to Text
- Automatic Speech Recognition
---
## Model Description
These are model weights originally provided by the authors of the paper [Learning Audio-Visual Speech Representation by Masked Multimodal Cluster Prediction](https://arxiv.org/pdf/2201.02184.pdf).
Video recordings of speech contain correlated audio and visual information, providing a strong signal for speech representation learning from the speaker’s lip
movements and the produced sound. Audio-Visual Hidden Unit BERT (AV-HuBERT), a self-supervised representation learning framework for
audio-visual speech, which masks multi-stream video input and predicts automatically discovered and iteratively refined multimodal hidden units. AV-HuBERT
learns powerful audio-visual speech representation benefiting both lip-reading and automatic speech recognition.
## Datasets
The authors trained the model on lip-reading benchmark LRS3 datasets (433 hours). |