Audio-visual Speech Enhancement Using Conditional Variational Auto-Encoders

Abstract

Variational auto-encoders (VAEs) are deep generative latent variable models that can be used for learning the distribution of complex data. VAEs have been successfully used to learn a probabilistic prior over speech signals, which is then used to perform speech enhancement. One advantage of this generative approach is that it does not require pairs of clean and noisy speech signals at training. In this paper, we propose audio-visual variants of VAEs for single-channel and speaker-independent speech enhancement. We develop a conditional VAE (CVAE) where the audio speech generative process is conditioned on visual information of the lip region. At test time, the audio-visual speech generative model is combined with a noise model based on nonnegative matrix factorization, and speech enhancement relies on a Monte Carlo expectation-maximization algorithm. Experiments are conducted with the recently published NTCD-TIMIT dataset as well as the GRID corpus. The results confirm that the proposed audio-visual CVAE effectively fuses audio and visual information, and it improves the speech enhancement performance compared with the audio-only VAE model, especially when the speech signal is highly corrupted by noise. We also show that the proposed unsupervised audio-visual speech enhancement approach outperforms a state-of-the-art supervised deep learning method.

Submitted to IEEE/ACM Transactions on Audio, Speech, and Language Processing

References in corpus (10)

• Adam: A Method for Stochastic Optimization
• Variational Inference: A Review for Statisticians
• Semi-Supervised Learning with Deep Generative Models
• Supervised and Unsupervised Speech Enhancement Using Nonnegative Matrix Factorization
• Statistical Speech Enhancement Based on Probabilistic Integration of Variational Autoencoder and Non-Negative Matrix Factorization
• Semi-supervised multichannel speech enhancement with variational autoencoders and non-negative matrix factorization
• A variance modeling framework based on variational autoencoders for speech enhancement
• DNN driven Speaker Independent Audio-Visual Mask Estimation for Speech Separation
• Speech enhancement with variational autoencoders and alpha-stable distributions
• A Deep Generative Model of Speech Complex Spectrograms

Cited by in corpus (13)

• Neural Target Speech Extraction: An Overview
• A survey of multimodal deep generative models
• Mixture of Inference Networks for VAE-based Audio-visual Speech Enhancement
• Integrating Uncertainty into Neural Network-based Speech Enhancement
• Robust Speaker Recognition Using Speech Enhancement And Attention Model
• Improved Lite Audio-Visual Speech Enhancement
• Expression-preserving face frontalization improves visually assisted speech processing
• Recent Advances and Challenges in Deep Audio-Visual Correlation Learning
• Face Frontalization Based on Robustly Fitting a Deformable Shape Model to 3D Landmarks
• Visual-Informed Speech Enhancement Using Attention-Based Beamforming
• Speaker Re-identification with Speaker Dependent Speech Enhancement
• End-to-end audio-visual learning for cochlear implant sound coding simulations in noisy environments
• SINVAD: Search-based Image Space Navigation for DNN Image Classifier Test Input Generation