Environment Classification via Blind Roomprints Estimation

09/15/2022

∙

In this paper we present a novel approach for environment classification for speech recordings, which does not require the selection of decaying reverberation tails. It is based on a multi-band RT60 analysis of blind channel estimates and achieves an accuracy of up to 93.6 from the ACE corpus.

READ FULL TEXT

Environment Classification via Blind Roomprints Estimation

Blind Room Parameter Estimation Using Multiple-Multichannel Speech Recordings

Blind Estimation of Room Acoustic Parameters and Speech Transmission Index using MTF-based CNNs

Realistic sources, receivers and walls improve the generalisability of virtually-supervised blind acoustic parameter estimators

Blind Deconvolution Microscopy Using Cycle Consistent CNN with Explicit PSF Layer

Blind Image Deblurring: a Review

Bias and variance reduction and denoising for CTF Estimation

Blind Judgement: Agent-Based Supreme Court Modelling With GPT

Environment Classification via Blind Roomprints Estimation

Related Research

Blind Room Parameter Estimation Using Multiple-Multichannel Speech Recordings

Blind Estimation of Room Acoustic Parameters and Speech Transmission Index using MTF-based CNNs

Realistic sources, receivers and walls improve the generalisability of virtually-supervised blind acoustic parameter estimators

Blind Deconvolution Microscopy Using Cycle Consistent CNN with Explicit PSF Layer

Blind Image Deblurring: a Review

Bias and variance reduction and denoising for CTF Estimation

Blind Judgement: Agent-Based Supreme Court Modelling With GPT