Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Whisper-Flamingo: Integrating Visual Features into Whisper for Audio-Visual Speech Recognition and Translation

Andrew Rouditchenko, Yuan Gong, Samuel Thomas, Leonid Karlinsky, Hilde Kuehne, Rogerio Feris, James Glass

TL;DR

Inspired by Flamingo which injects visual features into language models, Whisper-Flamingo is proposed which integrates visual features into the Whisper speech recognition and translation model with gated cross attention and outperforms audio-only Whisper on English speech recognition and En-X translation for 6 languages in noisy conditions.

Abstract

Audio-Visual Speech Recognition (AVSR) uses lip-based video to improve performance in noise. Since videos are harder to obtain than audio, the video training data of AVSR models is usually limited to a few thousand hours. In contrast, speech models such as Whisper are trained with hundreds of thousands of hours of data, and thus learn a better speech-to-text decoder. The huge training data difference motivates us to adapt Whisper to handle video inputs. Inspired by Flamingo which injects visual features into language models, we propose Whisper-Flamingo which integrates visual features into the Whisper speech recognition and translation model with gated cross attention. Our models achieve state-of-the-art ASR WER (0.68%) and AVSR WER (0.76%) on LRS3, and state-of-the-art ASR WER (1.3%) and AVSR WER (1.4%) on LRS2. Audio-visual Whisper-Flamingo outperforms audio-only Whisper on English speech recognition and En-X translation for 6 languages in noisy conditions. Moreover, Whisper-Flamingo is versatile and conducts all of these tasks using one set of parameters, while prior methods are trained separately on each language.

Whisper-Flamingo: Integrating Visual Features into Whisper for Audio-Visual Speech Recognition and Translation

TL;DR

Inspired by Flamingo which injects visual features into language models, Whisper-Flamingo is proposed which integrates visual features into the Whisper speech recognition and translation model with gated cross attention and outperforms audio-only Whisper on English speech recognition and En-X translation for 6 languages in noisy conditions.

Abstract

Audio-Visual Speech Recognition (AVSR) uses lip-based video to improve performance in noise. Since videos are harder to obtain than audio, the video training data of AVSR models is usually limited to a few thousand hours. In contrast, speech models such as Whisper are trained with hundreds of thousands of hours of data, and thus learn a better speech-to-text decoder. The huge training data difference motivates us to adapt Whisper to handle video inputs. Inspired by Flamingo which injects visual features into language models, we propose Whisper-Flamingo which integrates visual features into the Whisper speech recognition and translation model with gated cross attention. Our models achieve state-of-the-art ASR WER (0.68%) and AVSR WER (0.76%) on LRS3, and state-of-the-art ASR WER (1.3%) and AVSR WER (1.4%) on LRS2. Audio-visual Whisper-Flamingo outperforms audio-only Whisper on English speech recognition and En-X translation for 6 languages in noisy conditions. Moreover, Whisper-Flamingo is versatile and conducts all of these tasks using one set of parameters, while prior methods are trained separately on each language.
Paper Structure (16 sections, 1 equation, 1 figure, 10 tables)

This paper contains 16 sections, 1 equation, 1 figure, 10 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Experiments on LRS3
  4. Experimental Setup
  5. Modality Fusion Ablation with Whisper-Medium
  6. Whisper-Flamingo English Speech Recognition
  7. Whisper-Flamingo En-X Speech Translation
  8. Experiments on LRS2
  9. Conclusion
  10. Acknowledgments
  11. Appendix
  12. Original Results Table
  13. LRS3: Testing Different Noise Types and Noise Levels
  14. LRS2: Testing Different Noise Types and Noise Levels
  15. Analysis of Gated Cross Attention Position
  16. ...and 1 more sections

Figures (1)

  • Figure 1: Diagram of Whisper-Flamingo based on Whisper radford2023robust and Flamingo alayrac2022flamingo. We first fine-tune all of Whisper's parameters using English audio for English transcription and En-X translation. To train Whisper-Flamingo, we freeze the audio model, add gated cross attention layers into Whisper's decoder attending to visual features from AV-HuBERT, and train the model on audio-visual inputs.