Kodiertes Audiosignal für Mehrkanal-Audiosignal

Details

Veröffentlichungs-Nr.

EP4800688

Anmeldetag

27. Februar 2025

Veröffentlichung

2. September 2026

Rechtsraum

EP

IPC

G10L19/008, G10L25/30

Abstract

An audio apparatus comprises a downmixer (105) generating a downmix audio signal by downmixing a multichannel audio signal. A trained artificial neural network (111) determines scale factors for time frequency intervals of the multichannel audio signal from samples of the multichannel audio signal or the downmix audio signal. A signal generator (113) generates a modified multichannel audio signal by applying the first scale factors to samples of the time frequency intervals and a parameter circuit (115) generates spatial parameters representing relative properties for channels of the first modified multi-channel audio signal. An output circuit (109) generates am encoded audio signal to comprise the downmix audio signal and the spatial parameters. The encoded audio signal may also include a residual signal. A complementary apparatus may receive the encoded audio signal and comprise a trained artificial neural network (111) which from the downmix audio signal can generate corresponding scale factors. An output multichannel audio signal is generated by upmixing the downmix audio signal using these scale factors and the received spatial parameters.

Anmelder

Firma

Koninklijke Philips N.V.

Land

🇳🇱 Niederlande

Fachgebiete