l
HiFi-GAN: High-Fidelity Denoising and Dereverberation
Based on Speech Deep Features in Adversarial Networks

Jiaqi Su, Zeyu Jin, Adam Finkelstein

[Paper]

Real-world audio recordings are often degraded by factors such as noise, reverberation, and equalization distortion. This paper introduces HiFi-GAN, a deep learning method to transform recorded speech to sound as though it had been recorded in a studio. We use an end-to-end feed-forward WaveNet architecture, trained with multi-scale adversarial discriminators in both the time domain and the time-frequency domain. It relies on the deep feature matching losses of the discriminators to improve the perceptual quality of enhanced speech. The proposed model generalizes well to new speakers, new speech content, and new environments. It significantly outperforms state-of-the-art baseline methods in both objective and subjective experiments.
Here, generator G includes a feed-forward WaveNet for speech enhancement, followed by a convolutional Postnet for cleanup. Discriminators evaluate the resulting waveform ($$D_W$$, at multiple resolutions) and mel-spectrogram ($$D_S$$).

Real Demo for Ted Talk

Original input:

HiFi-GAN enhanced result:

Real Demo for VCTK Noisy

Original input:

HiFi-GAN enhanced result:

Real Demo for DAPS

Original input:

HiFi-GAN enhanced result:

* Using a model trained on our augmented synthetic dataset with speech corpus from the DAPS Dataset [7] and room impulse responses from MIT IR Survey Dataset [8].

SAMPLES