Sorry, you need to enable JavaScript to visit this website.

Language and Noise Transfer in Speech Enhancement Generative Adversarial Network

Citation Author(s):
Maruchan Park, Joan Serrà, Antonio Bonafonte, Kang-Hun Ahn
Submitted by:
Santi Pascual
Last updated:
19 April 2018 - 4:44pm
Document Type:
Presentation Slides
Document Year:
2018
Event:
Presenters:
Santiago Pascual
Paper Code:
4272
 

Speech enhancement deep learning systems usually require large amounts of training data to operate in broad conditions or real applications. This makes the adaptability of those systems into new, low resource environments an important topic. In this work, we present the results of adapting a speech enhancement generative adversarial network by fine-tuning the generator with small amounts of data. We investigate the minimum requirements to obtain a stable behavior in terms of several objective metrics in two very different languages: Catalan and Korean. We also study the variability of test performance to unseen noise as a function of the amount of different types of noise available for training. Results show that adapting a pre-trained English model with 10\,min of data already achieves a comparable performance to having two orders of magnitude more data. They also demonstrate the relative stability in test performance with respect to the number of training noise types.

up
0 users have voted: