Deep Audio-Visual Speech Recognition

TL;DR

This paper advances lip reading by comparing transformer-based models with different loss functions, exploring audio-visual integration, and introducing a new dataset, achieving state-of-the-art results in unconstrained speech recognition.

Contribution

It presents a comprehensive comparison of transformer-based lip reading models, investigates their complementarity with audio recognition, and releases a new challenging dataset for open-world speech recognition.

Findings

Transformer models outperform previous lip reading methods.

Lip reading significantly complements noisy audio speech recognition.

The new dataset enables robust evaluation of audio-visual speech models.

Abstract

The goal of this work is to recognise phrases and sentences being spoken by a talking face, with or without the audio. Unlike previous works that have focussed on recognising a limited number of words or phrases, we tackle lip reading as an open-world problem - unconstrained natural language sentences, and in the wild videos. Our key contributions are: (1) we compare two models for lip reading, one using a CTC loss, and the other using a sequence-to-sequence loss. Both models are built on top of the transformer self-attention architecture; (2) we investigate to what extent lip reading is complementary to audio speech recognition, especially when the audio signal is noisy; (3) we introduce and publicly release a new dataset for audio-visual speech recognition, LRS2-BBC, consisting of thousands of natural sentences from British television. The models that we train surpass the performance of all previous work on a lip reading benchmark dataset by a significant margin.