US 12,413,906 B2
Audio signal enhancement method, apparatus, device, and readable storage medium
Yangzhen Chen, Nanjing (CN); and Lijian Ye, Nanjing (CN)
Assigned to AAC Technologies (Nanjing) Co., Ltd., Nanjing (CN)
Filed by AAC Technologies (Nanjing) Co., Ltd., Nanjing (CN)
Filed on May 31, 2023, as Appl. No. 18/327,009.
Application 18/327,009 is a continuation of application No. PCT/CN2023/081940, filed on Mar. 16, 2023.
Claims priority of application No. 202211649357.3 (CN), filed on Dec. 21, 2022.
Prior Publication US 2024/0214730 A1, Jun. 27, 2024
Int. Cl. H04R 3/04 (2006.01); G06F 18/241 (2023.01)
CPC H04R 3/04 (2013.01) [G06F 18/241 (2023.01); H04R 2430/01 (2013.01)] 7 Claims
OG exemplary drawing
 
1. An audio signal enhancement method, comprising:
obtaining a first audio feature corresponding to an actual audio signal;
inputting the first audio feature to a trained classifier for classification and identification, to obtain audio-type representation data corresponding to the actual audio signal; and
enhancing a target audio signal conforming to a target audio type in the actual audio signal with reference to the audio-type representation data, to obtain an enhanced audio signal;
wherein the step of enhancing the target audio signal conforming to the target audio type in the actual audio signal with reference to the audio-type representation data to obtain the enhanced audio signal comprises:
performing a median filtering on the audio-type representation data for a predetermined number of times to obtain audio-type representation data without outliers; and
performing a gaining and/or a dynamic range enhancement on the target audio signal in different frequency bands conforming to the target audio type in the actual audio signal when the audio-type representation data without outliers correspond to the target audio type, to obtain the enhanced audio signal;
wherein the step of performing the gaining and/or the dynamic range enhancement on the target audio signal in the actual audio signal in different frequency bands conforming to the target audio type comprises:
performing the gaining with reference to a predetermined equalizer fade-in and fade-out time, and/or, performing the dynamic range enhancement with reference to a predetermined time parameter for dynamic range control, on the target audio signal in the different frequency bands conforming to the target audio type in the actual audio signal;
wherein after the step of enhancing the target audio signal conforming to the target audio type in the actual audio signal with reference to the audio-type representation data to obtain the enhanced audio signal, the method further comprises:
performing an amplitude limiting processing on the enhanced audio signal to obtain a clipping-free enhanced audio signal.