Multi-channel Late Reverberation Power Spectral Density Estimation Based on Nuclear Norm Minimization
Reverberant acoustic systems with a single-source and 4 microphones
Acoustic system 1: T60 = 0.61 s, source at θ = 45°
Acoustic system 2: T60 = 1.25 s, source at θ = -65°
WOLA framework with a frame size of 64 ms and an overlap of 75%
Dereverberation using the true RETFs for acoustic system 1 (θ = 45°)
Dereverberation using the true RETFs for acoustic system 2 (θ = -65°)
Dereverberation using erroneous RETFs for acoustic system 1 (θ = 45°)
This page is maintained by
Ina Kodrasi
Last modification: