Personalized bandwidth extension
AssigneeQUALCOMM Incorporated
InventorsLae-Hoon Kim, Sang-Uk Ryu, Jongwon Shin
Life of this patent
Estimating the term…
Abstract
A personalized (i.e., speaker-derivable) bandwidth extension is provided in which the model used for bandwidth extension is personalized (e.g., tailored) to each specific user. A training phase is performed to generate a bandwidth extension model that is personalized to a user. The model may be subsequently used in a bandwidth extension phase during a phone call involving the user. The bandwidth extension phase, using the personalized bandwidth extension model, will be activated when a higher band (e.g., wideband) is not available and the call is taking place on a lower band (e.g., narrowband).
Claims44
A method for bandwidth extension, comprising:
- generating a bandwidth extension model with data of an identified speaker, wherein the bandwidth extension model is only generated if an acoustic condition of the speaker's voice is of high enough quality to be used for training;
- and
- performing bandwidth extension using the bandwidth extension model on a narrowband call including the identified speaker.
An apparatus for bandwidth extension, comprising:
- means for generating a bandwidth extension model with data of an identified speaker, wherein the bandwidth extension model is only generated if an acoustic condition of the speaker's voice is of high enough quality to be used for training;
- and
- means for performing bandwidth extension using the bandwidth extension model on a narrowband call including the identified speaker.
A non-transitory computer-readable medium comprising instructions that cause a computer to:
- generate a bandwidth extension model with data of an identified speaker, wherein the bandwidth extension model is only generated if an acoustic condition of the speaker's voice is of high enough quality to be used for training;
- and
- perform bandwidth extension using the bandwidth extension model on a narrowband call including the identified speaker.
A system for bandwidth extension, comprising:
- a training module adapted to generate a bandwidth extension model with data of an identified speaker, wherein the bandwidth extension model is only generated if an acoustic condition of the speaker's voice is of high enough quality to be used for training;
- and
- a processor for performing bandwidth extension using the bandwidth extension model on a narrowband call including the identified speaker.
Dependencies are inferred from the claim text.