We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.AS

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Audio and Speech Processing

Title: Ensemble of Discriminators for Domain Adaptation in Multiple Sound Source 2D Localization

Abstract: This paper introduces an ensemble of discriminators that improves the accuracy of a domain adaptation technique for the localization of multiple sound sources. Recently, deep neural networks have led to promising results for this task, yet they require a large amount of labeled data for training. Recording and labeling such datasets is very costly, especially because data needs to be diverse enough to cover different acoustic conditions. In this paper, we leverage acoustic simulators to inexpensively generate labeled training samples. However, models trained on synthetic data tend to perform poorly with real-world recordings due to the domain mismatch. For this, we explore two domain adaptation methods using adversarial learning for sound source localization which use labeled synthetic data and unlabeled real data. We propose a novel ensemble approach that combines discriminators applied at different feature levels of the localization model. Experiments show that our ensemble discrimination method significantly improves the localization performance without requiring any label from the real data.
Comments: Please refer to final version at arXiv:2012.05533
Subjects: Audio and Speech Processing (eess.AS); Machine Learning (cs.LG); Sound (cs.SD)
Cite as: arXiv:2012.05908 [eess.AS]
  (or arXiv:2012.05908v2 [eess.AS] for this version)

Submission history

From: Guillaume Le Moing [view email]
[v1] Thu, 10 Dec 2020 09:17:29 GMT (1728kb)
[v2] Tue, 16 Mar 2021 08:42:17 GMT (1728kb)

Link back to: arXiv, form interface, contact.