End-To-End Label Uncertainty Modeling for Speech-based Arousal Recognition Using Bayesian Neural Networks

1 code implementation • 7 Oct 2021 • Navin Raj Prabhu, Guillaume Carbajal, Nale Lehmann-Willenbrock, Timo Gerkmann

At training, the network learns a distribution of weights to capture the inherent uncertainty related to subjective arousal annotations.

Speech Emotion Recognition

Disentanglement Learning for Variational Autoencoders Applied to Audio-Visual Speech Enhancement

1 code implementation • 19 May 2021 • Guillaume Carbajal, Julius Richter, Timo Gerkmann

In this work, we propose to use an adversarial training scheme for variational autoencoders to disentangle the label from the other latent variables.

Attribute Decoder +2

Variational Autoencoder for Speech Enhancement with a Noise-Aware Encoder

no code implementations • 17 Feb 2021 • Huajian Fang, Guillaume Carbajal, Stefan Wermter, Timo Gerkmann

Recently, a generative variational autoencoder (VAE) has been proposed for speech enhancement to model speech statistics.

Speech Enhancement

Guided Variational Autoencoder for Speech Enhancement With a Supervised Classifier

no code implementations • 12 Feb 2021 • Guillaume Carbajal, Julius Richter, Timo Gerkmann

In this paper, we propose to guide the variational autoencoder with a supervised classifier separately trained on noisy speech.

Speech Enhancement

Joint NN-Supported Multichannel Reduction of Acoustic Echo, Reverberation and Noise

no code implementations • 20 Nov 2019 • Guillaume Carbajal, Romain Serizel, Emmanuel Vincent, Eric Humbert

We consider the problem of simultaneous reduction of acoustic echo, reverberation and noise.