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

Download:

Current browse context:

cs.NE

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

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

Computer Science > Neural and Evolutionary Computing

Title: Mean absorption estimation from room impulse responses using virtually supervised learning

Authors: Cédric Foy (UMRAE ), Antoine Deleforge (MULTISPEECH), Diego Di Carlo (PANAMA)
Abstract: In the context of building acoustics and the acoustic diagnosis of an existing room, this paper introduces and investigates a new approach to estimate mean absorption coefficients solely from a room impulse response (RIR). This inverse problem is tackled via virtually-supervised learning, namely, the RIR-to-absorption mapping is implicitly learned by regression on a simulated dataset using artificial neural networks. We focus on simple models based on well-understood architectures. The critical choices of geometric, acoustic and simulation parameters used to train the models are extensively discussed and studied, while keeping in mind conditions that are representative of the field of building acoustics. Estimation errors from the learned neural models are compared to those obtained with classical formulas that require knowledge of the room's geometry and reverberation times. Extensive comparisons made on a variety of simulated test sets highlight different conditions under which the learned models can overcome the well-known limitations of the diffuse sound field hypothesis underlying these formulas. Results obtained on real RIRs measured in an acoustically configurable room show that at 1~kHz and above, the proposed approach performs comparably to classical models when reverberation times can be reliably estimated, and continues to work even when they cannot.
Subjects: Neural and Evolutionary Computing (cs.NE); Sound (cs.SD); Audio and Speech Processing (eess.AS); Classical Physics (physics.class-ph)
Journal reference: Journal of the Acoustical Society of America, Acoustical Society of America, 2021, 150 (2), pp.1286-1299
DOI: 10.1121/10.0005888
Cite as: arXiv:2109.00393 [cs.NE]
  (or arXiv:2109.00393v1 [cs.NE] for this version)

Submission history

From: Antoine Deleforge [view email]
[v1] Wed, 1 Sep 2021 14:06:20 GMT (1216kb,D)

Link back to: arXiv, form interface, contact.