Personalized real-time audio generation based on user physiological response
Systems and methods are provided for personalized real-time audio generation based on user physiological response. An example method includes obtaining hearing information associated with a user's hearing capabilities, the hearing information indicating one or more constraints on the user's hearing, and the hearing information being determined based on one or more hearing tests performed by the user; requesting listening history information from one or more music streaming platforms, the listening history information indicating, at least, music playlists to which a user created or is subscribed; determining, based on the listening history information, a style preference associated with the user; generating, utilizing one or more machine learning models, personalized music based on the hearing information and the style preference, wherein the personalized music comports with the constraints, and wherein the system is configured to provide the personalized music for output via a user device of the user.
1. A computer-implemented method comprising:
by a system of one or more computers,
obtaining hearing information associated with a user's hearing capabilities, the hearing information indicating one or more constraints, and the hearing information being determined based on one or more hearing tests performed by the user;
receiving listening history information from one or more music streaming platforms, the listening history information indicating, at least, music playlists to which a user created or is subscribed;
determining, based on the listening history information, a style preference of the user, the style preference encoding style information associated with the listening history information; and
generating, utilizing one or more machine learning models, personalized music which expands upon a musical cue in accordance with the style preference, the personalized music being generated using the hearing information, wherein the personalized music comports with the constraints, and wherein the system is configured to provide the personalized music for output via a user device of the user.
2. The computer-implemented method of claim 1 , wherein the hearing information indicates that the user has tinnitus, or wherein the hearing information indicates a frequency range associated with reduced hearing capability.
3. The computer-implemented method of claim 2 , wherein the hearing information indicates that the user has tinnitus, and wherein generating personalized music comprises:
identifying, from the hearing information, a tone associated with the user's tinnitus; and
generating personalized music, such that the one or more machine learning models set the identified tone as a musical note for utilization in the personalized music.
4. The computer-implemented method of claim 3 , wherein remaining musical notes utilized in the personalized music are associated with respective intervals from the identified tone.
5. The computer-implemented method of claim 2 , wherein the hearing information indicates a frequency range, and wherein generating personalized music comprises:
selecting, via the one or more machine learning models, instruments associated with musical notes outside of the frequency range; and
generating personalized music utilizing the selected instruments.
6. The computer-implemented method of claim 1 , wherein receiving listening history information comprises:
receiving, from a user device executing an electronic game, information indicating initiation of the electronic game;
obtaining authentication information associated with the user; and
requesting the listening history information utilizing the authentication information.
7. The computer-implemented method of claim 1 , wherein determining the style preference comprises:
grouping a threshold number of songs identified in the listening history information; and
determining a data structure representing the style preference, the data structure comprising assigned weights to each of a plurality of musical styles,
wherein the data structure is provided as an input to the one or more machine learning models.
8. The computer-implemented method of claim 1 , wherein the one or more machine learning models comprise a recurrent neural network or a convolutional neural network.
9. The computer-implemented method of claim 1 , wherein the musical cue represents a portion of audio associated with the style preference or a musical score associated with the style preference.
10. A system comprising one or more computers and computer-storage media storing instructions that when executed by the one or more computers, cause the one or more computers to perform operations comprising:
obtaining hearing information associated with a user's hearing capabilities, the hearing information indicating one or more constraints, and the hearing information being determined based on one or more hearing tests performed by the user;
receiving listening history information from one or more music streaming platforms, the listening history information indicating, at least, music playlists to which a user created or is subscribed;
determining, based on the listening history information, a style preference of the user, the style preference encoding style information associated with the listening history information; and
generating, utilizing one or more machine learning models, personalized music which expands upon a musical cue in accordance with the style preference, the personalized music being generated using the hearing information, wherein the personalized music comports with the constraints, and wherein the system is configured to provide the personalized music for output via a user device of the user.
11. The system of claim 10 , wherein the hearing information indicates that the user has tinnitus, or wherein the hearing information indicates a frequency range associated with reduced hearing capability.
12. The system of claim 11 , wherein the hearing information indicates that the user has tinnitus, and wherein generating personalized music comprises:
identifying, from the hearing information, a tone associated with the user's tinnitus; and
generating personalized music, such that the one or more machine learning models set the identified tone as a musical note for utilization in the personalized music.
13. The system of claim 12 , wherein remaining musical notes utilized in the personalized music are associated with respective intervals from the identified tone.
14. The system of claim 10 , wherein the hearing information indicates a frequency range, and wherein generating personalized music comprises:
selecting, via the one or more machine learning models, instruments associated with musical notes outside of the frequency range; and
generating personalized music utilizing the selected instruments.
15. The system of claim 10 , wherein the one or more machine learning models comprise a recurrent neural network or a convolutional neural network.
16. Non-transitory computer storage media storing instructions that when executed by a system of one or more computers, cause the one or more computers to perform operations comprising:
obtaining hearing information associated with a user's hearing capabilities, the hearing information indicating one or more constraints, and the hearing information being determined based on one or more hearing tests performed by the user;
receiving listening history information from one or more music streaming platforms, the listening history information indicating, at least, music playlists to which a user created or is subscribed;
determining, based on the listening history information, a style preference of the user, the style preference encoding style information associated with the listening history information; and
generating, utilizing one or more machine learning models, personalized music which expands upon a musical cue in accordance with the style preference, the personalized music being generated using the hearing information, wherein the personalized music comports with the constraints, and wherein the system is configured to provide the personalized music for output via a user device of the user.
17. The computer storage media of claim 16 , wherein the hearing information indicates that the user has tinnitus, or wherein the hearing information indicates a frequency range associated with reduced hearing capability.
18. The computer storage media of claim 17 , wherein the hearing information indicates that the user has tinnitus, and wherein generating personalized music comprises:
identifying, from the hearing information, a tone associated with the user's tinnitus; and
generating personalized music, such that the one or more machine learning models set the identified tone as a musical note for utilization in the personalized music.
19. The computer storage media of claim 17 , wherein the hearing information indicates a frequency range, and wherein generating personalized music comprises:
selecting, via the one or more machine learning models, instruments associated with musical notes outside of the frequency range; and
generating personalized music utilizing the selected instruments.
20. The computer storage media of claim 16 , wherein the one or more machine learning models comprise a recurrent neural network or a convolutional neural network.