# US 2022/0044693 A1 · Internet calling method and apparatus, computer device, and storage medium > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2022/0044693 A1 - Document ID: US20220044693A1 - Status: Published application - Kind: A1 - Application number: 17/507,533 - Filed: 2021-10-21 - Published: 2022-02-10 - Earliest claimed: 2019-12-10 - Assignees: Tencent Technology (Shenzhen) Company Limited - Applicants: Tencent Technology (Shenzhen) Company Limited - Inventors: Junbin LIANG - Web page: https://patentgrep.com/patent/US20220044693A1 ## CPC - H04L 65/80: Responding to QoS - G10L 19/02: using spectral analysis, e.g. transform vocoders or subband vocoders - G10L 19/06: Determination or coding of the spectral characteristics, e.g. of the short-term prediction coefficients - G10L 19/167: {Audio streaming, i.e. formatting and decoding of an encoded audio signal representation into a data stream for transmission or storage purposes} - G10L 19/24: Variable rate codecs, e.g. for generating different qualities using a scalable representation such as hierarchical encoding or layered encoding - H04B 17/3913: {Predictive models, e.g. based on neural network models} - H04L 65/1083: In-session procedures - H04L 65/70: Media network packetisation - H04L 65/75: Media network packet handling - H04L 65/752: adapting media to network capabilities - G10L 19/032: Quantisation or dequantisation of spectral components - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - H04B: TRANSMISSION ## Abstract This disclosure provides a network call method and apparatus, a computer device, and a storage medium, and belongs to the field of audio data processing. The method includes: performing time-frequency transformation on an acquired audio signal, to obtain a plurality of pieces of frequency domain information of the audio signal; determining a target bit rate corresponding to the audio signal according to the plurality of pieces of frequency domain information; and encoding the audio signal based on the target bit rate, and performing a network call based on the encoded audio signal. ## Claims ### Claim 1 A network call method, applicable to a computer device, the method comprising: performing time-frequency transformation on an acquired audio signal to obtain a plurality of pieces of frequency domain information of the audio signal; determining a target bit rate corresponding to the audio signal according to the plurality of pieces of frequency domain information; encoding the audio signal based on the target bit rate; and performing a network call based on the encoded audio signal. ### Claim 2 (depends on 1) The method according to claim 1, wherein determining the target bit rate corresponding to the audio signal according to the plurality of pieces of frequency domain information comprises: inputting the plurality of pieces of frequency domain information into a bit rate prediction model to obtain a plurality of first candidate bit rates, the plurality of first candidate bit rates being sufficient to meet a target speech quality condition; and determining a first bit rate that meets a target condition from the plurality of first candidate bit rates as the target bit rate. ### Claim 3 (depends on 2) The method according to claim 2, wherein determining the first bit rate that meets the target condition from the plurality of first candidate bit rates as the target bit rate comprises: classifying the plurality of first candidate bit rates having same bit rate values into bit rate groups; obtaining a quantity of each bit rate group; and determining a bit rate of a bit rate group having the largest quantity as the target bit rate. ### Claim 4 (depends on 2) The method according to claim 2, further comprising: performing model training based on a sample data set to obtain the bit rate prediction model, the sample data set comprising a plurality of sample audio signals and sample bit rates corresponding to the plurality of sample audio signals, the plurality of sample audio signals meeting a same speech quality condition. ### Claim 5 (depends on 4) The method according to claim 4, wherein before performing model training based on the sample data set to obtain the bit rate prediction model, the method further comprises: encoding and then decoding the plurality of sample audio signals based on the sample bit rates corresponding to the plurality of sample audio signals to obtain a plurality of decoded sample audio signals; and determining speech qualities of the plurality of sample audio signals based on the plurality of sample audio signals and the plurality of decoded sample audio signals. ### Claim 6 (depends on 4) The method according to claim 4, wherein performing model training based on the sample data set to obtain the bit rate prediction model comprises: performing time-frequency transformation on the plurality of sample audio signals to obtain a plurality of pieces of sample frequency domain information; inputting the plurality of pieces of sample frequency domain information into an initial model; performing, by the initial model, bit rate prediction based on the plurality of pieces of sample frequency domain information to obtain a plurality of predicted bit rates corresponding to the plurality of pieces of sample frequency domain information; iteratively adjust one or more model parameters of the initial model based on differential information between the sample bit rates and the predicted bit rates until the a target cut-off condition is met; and using the initial model with the adjusted one or more model parameters as the bit rate prediction model. ### Claim 7 (depends on 1) The method according to claim 1, wherein performing time-frequency transformation on the acquired audio signal to obtain the plurality of pieces of frequency domain information comprises: dividing the audio signal into a plurality of audio frames, two audio frames adjacent in time having an overlapping part; performing windowing on the audio frames to obtain a plurality of windowed audio frames; and performing Fourier transformation on the plurality of windowed audio frames to obtain the plurality of pieces of frequency domain information. ### Claim 8 (depends on 7) The method according to claim 7, wherein performing Fourier transformation on the plurality of windowed audio frames to obtain the plurality of pieces of frequency domain information comprises: performing Fourier transformation on the plurality of windowed audio frames to obtain linear frequency-domain parameters of the audio frames; obtaining linear frequency-domain-power-spectrum values of the audio frames based on the linear frequency-domain parameters of the audio frames; obtaining a plurality of Bark-domain-power-spectrum values based on the linear frequency-domain-power-spectrum values of the audio frames; and performing a logarithmic operation on the plurality of Bark-domain-power-spectrum values to obtain the plurality of pieces of frequency domain information. ### Claim 9 (depends on 1) The method according to claim 1, wherein performing time-frequency transformation on an acquired audio signal to obtain the plurality of pieces of frequency domain information of the audio signal comprises: extracting a human voice signal from the audio signal, performing time-frequency transformation on the human voice signal to obtain the plurality of pieces of frequency domain information of the audio signal; or recognizing a start time point and an end time point of a human voice in the audio signal, and performing time-frequency transformation on an audio signal between the start time point and the end time point to obtain the plurality of pieces of frequency domain information of the audio signal. ### Claim 10 A computer device, comprising one or more processors and one or more memories, the one or more memories storing at least one program code, the at least one program code, when being loaded and executed by the one or more processors, causing the computer device to perform steps, comprising: performing time-frequency transformation on an acquired audio signal to obtain a plurality of pieces of frequency domain information of the audio signal; determining a target bit rate corresponding to the audio signal according to the plurality of pieces of frequency domain information; encoding the audio signal based on the target bit rate; and performing a network call based on the encoded audio signal. ### Claim 11 (depends on 10) The computer device according to claim 10, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: inputting the plurality of pieces of frequency domain information into a bit rate prediction model to obtain a plurality of first candidate bit rates, the plurality of first candidate bit rates being sufficient to meet a target speech quality condition; and determining a first bit rate that meets a target condition from the plurality of first bit rates as the target bit rate. ### Claim 12 (depends on 11) The computer device according to claim 11, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: classifying the plurality of first candidate bit rates having the same bit rate value into a bit rate group; obtaining a quantity of each bit rate group; and determining a bit rate of a bit rate group having the largest quality as the target bit rate. ### Claim 13 (depends on 11) The computer device according to claim 11, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: performing model training based on a sample data set to obtain the bit rate prediction model, the sample data set comprising a plurality of sample audio signals and sample bit rates corresponding to the plurality of sample audio signals, the plurality of sample audio signals meeting a same speech quality condition. ### Claim 14 (depends on 13) The computer device according to claim 13, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: encoding and then decoding the plurality of sample audio signals based on the sample bit rates corresponding to the plurality of sample audio signals to obtain a plurality of decoded sample audio signals; and determining speech qualities of the plurality of sample audio signals based on the plurality of sample audio signals and the plurality of decoded sample audio signals. ### Claim 15 (depends on 13) The computer device according to claim 13, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: performing time-frequency transformation on the plurality of sample audio signals to obtain a plurality of pieces of sample frequency domain information; inputting the plurality of pieces of sample frequency domain information into an initial model; performing, by the initial model, bit rate prediction based on the plurality of pieces of sample frequency domain information to obtain a plurality of predicted bit rates corresponding to the plurality of pieces of sample frequency domain information; iteratively adjust one or more model parameters of the initial model based on differential information between the sample bit rates and the predicted bit rates until a target cut-off condition is met; and using the initial model with the adjusted one or more model parameters as the bit rate prediction model. ### Claim 16 (depends on 10) The computer device according to claim 10, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: dividing the audio signal into a plurality of audio frames, two audio frames adjacent in time having an overlapping part; performing windowing on the audio frames to obtain a plurality of windowed audio frames; and performing Fourier transformation on the plurality of windowed audio frames to obtain the plurality of pieces of frequency domain information. ### Claim 17 (depends on 16) The computer device according to claim 16, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: performing Fourier transformation on the plurality of windowed audio frames to obtain linear frequency-domain parameters of the audio frames; obtaining linear frequency-domain-power-spectrum values of the audio frames based on the linear frequency-domain parameters of the audio frames; obtaining a plurality of Bark-domain-power-spectrum values based on the linear frequency-domain-power-spectrum values of the audio frames; and performing a logarithmic operation on the plurality of Bark-domain-power-spectrum values to obtain the plurality of pieces of frequency domain information. ### Claim 18 (depends on 10) The computer device according to claim 10, wherein the one or more processors are configured to cause the computer device to perform the steps comprising: extracting a human voice signal from the audio signal, performing time-frequency transformation on the human voice signal to obtain the plurality of pieces of frequency domain information of the audio signal; or recognizing a start time point and an end time point of a human voice in the audio signal, and performing time-frequency transformation on an audio signal between the start time point and the end time point to obtain the plurality of pieces of frequency domain information of the audio signal. ### Claim 19 A non-transitory computer-readable storage medium, storing at least one program code, the at least one program code being loaded and executed by a processor to cause a computer device to perform steps comprising: performing time-frequency transformation on an acquired audio signal, to obtain a plurality of pieces of frequency domain information of the audio signal; determining a target bit rate corresponding to the audio signal according to the plurality of pieces of frequency domain information; and encoding the audio signal based on the target bit rate; and performing a network call based on the encoded audio signal. ### Claim 20 (depends on 19) The non-transitory computer-readable storage medium of claim 19, wherein the at least one program code is further configured to be loaded and executed by the processor to cause the computer device to perform steps comprising: inputting the plurality of pieces of frequency domain information into a bit rate prediction model to obtain a plurality of first candidate bit rates, the plurality of first candidate bit rates being sufficient to meet a target speech quality condition; and determining a first bit rate that meets a target condition from the plurality of first candidate bit rates as the target bit rate.