Visual Speech Enhancement

TL;DR

This paper introduces an audio-visual neural network for speech enhancement that leverages visual mouth movements to improve speech clarity in noisy environments, outperforming previous methods on multiple datasets including real-world videos.

Contribution

The paper presents a novel audio-visual neural network trained with augmented data, demonstrating superior performance in speech enhancement and generalization to diverse noise conditions and datasets.

Findings

Outperforms prior audio-visual methods on lipreading datasets

Effective on real-world videos like Barack Obama's addresses

Generalizes well to different noise types

Abstract

When video is shot in noisy environment, the voice of a speaker seen in the video can be enhanced using the visible mouth movements, reducing background noise. While most existing methods use audio-only inputs, improved performance is obtained with our visual speech enhancement, based on an audio-visual neural network. We include in the training data videos to which we added the voice of the target speaker as background noise. Since the audio input is not sufficient to separate the voice of a speaker from his own voice, the trained model better exploits the visual input and generalizes well to different noise types. The proposed model outperforms prior audio visual methods on two public lipreading datasets. It is also the first to be demonstrated on a dataset not designed for lipreading, such as the weekly addresses of Barack Obama.