Apple Inc.

Aalto-korkeakoulusaatio

Symeon Delikaris-Manias

Joshua D Atkins

Ville Topias Pulkki

Mehrez Souden

Peter A Raffensperger

Impulse responses of a device are measured. A database of sound files is generated by convolving source signals with the impulse responses of the device. The sound files from the database are transformed into time-frequency domain. One or more sub-band directional features is estimated at each sub-band of the time-frequency domain. A deep neural network (DNN) is trained for each sub-band based on the estimated one or more sub-band directional features and a target directional feature.

Machine learning based sound field analysis

Method for spatial filtering of at least one sound signal (M; W(n)) includes the steps of: generation of a first, a second and a third captured sound signal by capturing of the respective sound signals by microphones characterized by directivity patterns of different orders; performing a short-time Fourier transformation of the captured, sound signal s; measuring a cross-pattern correlation or a cross-pattern coherence towards a desired direction ([phi]); calculation of a gain factor (G+) using a cross-pattern correlation based on time-averaged correlation or coherence between the first captured sound signal and the second captured sound signal; and applying the gain factor (G+) to the corresponding time-frequency positions in the third captured sound, signal (; M; W(n)). Independent patent claims also for a system and computer readable storage medium.

Method for spatial filtering of at least one sound signal, computer readable storage medium and spatial filtering system based on cross-pattern coherence

Growing community of inventors

Helsinki, Finland

Symeon Delikaris-Manias