CPC G10L 19/0017 (2013.01) [H04L 1/04 (2013.01); H04W 4/80 (2018.02); H04W 28/04 (2013.01)] | 16 Claims |
1. An audio data recovery method, comprising:
dividing audio data into a first frequency domain component and a second frequency domain component in a frequency domain, wherein said dividing audio data into a first frequency domain component and a second frequency domain component in a frequency domain comprises:
dividing the audio data into a first type of frame data and a second type of frame data in the frequency domain; and
dividing the first type of frame data into the first frequency domain component and the second frequency domain component, wherein said dividing the first type of frame data into the first frequency domain component and the second frequency domain component further comprises:
estimating a power spectrum of the first type of frame data;
determining peaks of the power spectrum;
determining candidate frequency domain components according to the peaks; and
regarding the candidate frequency domain components whose power is greater than a preset threshold as the second frequency domain component, and other candidate frequency domain components as the first frequency domain component;
using a second data recovery algorithm to recover the audio data in the second frequency domain component; and
using a first data recovery algorithm with less complexity than the second data recovery algorithm to recover the audio data in the first frequency domain component.
|