Audio processing method and apparatus
View Patent ↗Embodiments of the present application provide an audio processing method and an apparatus. The method includes: a mobile terminal and a vehicle terminal are in a connected state, and playing, by the mobile terminal, a first audio synchronously with the vehicle terminal; obtaining, by the mobile terminal, a recorded audio of a current environment, where the recorded audio includes the first audio played by the vehicle terminal and a second audio for voice recognition; and eliminating, according to the first audio played by the mobile terminal, the first audio played by the vehicle terminal in the recorded audio to obtain the second audio. In the embodiments of the present application, by playing the first audio synchronously by the mobile terminal and the vehicle terminal, the second audio for voice recognition in the recorded audio can be obtained according to the first audio played by the mobile terminal.
1. An audio processing method, wherein a mobile terminal and a vehicle terminal are in a connected state, the method is applied to the mobile terminal, and the method comprises: playing a first audio synchronously with the vehicle terminal, wherein an amplitude corresponding to the first audio when being played by the mobile terminal is 0; obtaining a recorded audio of a current environment, wherein the recorded audio comprises the first audio played by the vehicle terminal and a second audio for voice recognition; and eliminating, according to the first audio played by the mobile terminal, the first audio played by the vehicle terminal in the recorded audio to obtain the second audio; wherein the eliminating, according to the first audio played by the mobile terminal, the first audio played by the vehicle terminal in the recorded audio to obtain the second audio comprises: performing a resampling processing on the first audio played by the mobile terminal to obtain a third audio; performing a dual channel to single channel processing on the third audio to obtain a fourth audio, wherein the third audio is dual channel data, and the fourth audio is single channel data; and eliminating the first audio played by the vehicle terminal in the recorded audio to obtain the second audio by taking the fourth audio as a reference audio; wherein performing time calibration on the reference audio and the recorded audio comprises: obtaining a first duration from a time when the mobile terminal obtains the recorded audio to a time when a voice recognition module of the mobile terminal receives the recorded audio, and obtaining a second duration from a time when the mobile terminal obtains the recorded audio to a time when the voice recognition module receives the reference audio corresponding to the recorded audio; subtracting the second duration from the first duration to obtain a transmission delay duration; and subtracting the transmission delay duration from a first time to obtain a second time, and determining an audio received by the voice recognition module at the second time as the reference audio corresponding to the recorded audio, wherein the first time is a time when the voice recognition module receives the recorded audio.
2. The method according to claim 1 , wherein the method further comprises:
caching the first audio locally before playing the first audio synchronously with the vehicle terminal.
3. An audio processing apparatus, wherein the audio processing apparatus and a vehicle terminal are in a connected state, and the apparatus comprises: a processor coupled to a memory; the memory is configured to store a computer program; and the processor is configured to invoke the computer program stored in the memory, which, when executed by the processor, causes the processor to: play a first audio synchronously with the vehicle terminal, wherein an amplitude corresponding to the first audio when being played by the audio processing apparatus is 0; obtain a recorded audio of a current environment, wherein the recorded audio comprises the first audio played by the vehicle terminal and a second audio for voice recognition; and eliminate, according to the first audio played by the audio processing apparatus, the first audio played by the vehicle terminal in the recorded audio to obtain the second audio; wherein the computer program further causes the processor to: perform a resampling processing on the first audio played by the audio processing apparatus to obtain a third audio; perform a dual channel to single channel processing on the third audio to obtain a fourth audio, wherein the third audio is dual channel data, and the fourth audio is single channel data; and eliminate the first audio played by the vehicle terminal in the recorded audio to obtain the second audio by taking the fourth audio as a reference audio; wherein the computer program further causes the processor to: obtain a first duration from a time when the mobile terminal obtains the recorded audio to a time when a voice recognition module receives the recorded audio, and obtain a second duration from a time when the mobile terminal obtains the recorded audio to a time when the voice recognition module receives the reference audio corresponding to the recorded audio; subtract the second duration from the first duration to obtain a transmission delay duration; and subtract the transmission delay duration from a first time to obtain a second time, and determine an audio received by the voice recognition module at the second time as the reference audio corresponding to the recorded audio, wherein the first time is a time when the voice recognition module receives the recorded audio.
4. The apparatus according to claim 3 , the computer program further causes the processor to:
cache the first audio locally before playing the first audio synchronously with the vehicle terminal.
5. A nonvolatile memory, wherein the nonvolatile memory has stored thereon a program or an instruction, wherein the method of claim 1 is executed when the program or the instruction is operated on a computer.