Adaptive speech intelligibility control for speech privacy
AssigneeHewlett-Packard Development Company, L.P.
InventorsSunil Bharitkar, Wensen Liu, Madhu Sudan Athreya, Richard Sweet
Life of this patent
Estimating the term…
Abstract
In some examples, adaptive speech intelligibility control for speech privacy may include determining, based on background noise at a near-end of a speaker, a noise estimate associated with speech emitted from the speaker, and comparing, by using a specified factor, the noise estimate to a speech level estimate for the speech emitted from the speaker. Adaptive speech intelligibility control for speech privacy may further include determining, based on the comparison, a gain value to be applied to the speaker to produce the speech at a specified level to maintain on-axis intelligibility with respect to the speaker, and applying the gain value to the speaker.
Claims18
An adaptive speech intelligibility control for speech privacy apparatus comprising:
- a processor;
- and
- a memory storing machine readable instructions that when executed by the processor cause the processor to:
- determine, based on background noise at a near-end of a speaker, a noise estimate associated with speech emitted from the speaker;
- compare, by using a specified factor that includes utilizing a masking threshold below which presence of any speech is rendered inaudible, the noise estimate to a speech level estimate for the speech emitted from the speaker;
- determine, based on the comparison, a gain value to be applied to the speaker to produce the speech at a specified level to maintain on-axis intelligibility with respect to the speaker;
- determine, based on applying a non-linear acoustic model, a modulated ultrasound level estimation of far-end speech;
- and
- apply the gain value and the modulated ultrasound level estimation to the speaker.
A method for adaptive speech intelligibility control for speech privacy comprising:
- determining, by a processor, based on background noise at a near-end of a speaker, a noise estimate associated with speech emitted from the speaker;
- comparing, by using a specified factor that includes utilizing a masking threshold below which presence of any speech is rendered inaudible, the noise estimate to a speech level estimate for the speech emitted from the speaker;
- determining, by using a camera, a distance of a target listener from the speaker;
- and
- determining, based on the comparison and the distance of the target listener from the speaker, a gain value to be applied to the speaker to produce the speech at a specified level to maintain on-axis intelligibility with respect to the speaker;
- determining, based on applying a non-linear acoustic model, a modulated ultrasound level estimation of far-end speech;
- and
- applying the gain value and the modulated ultrasound level estimation to the speaker.
A non-transitory computer readable medium having stored thereon machine readable instructions to provide adaptive speech intelligibility control for speech privacy, the machine readable instructions, when executed, cause a processor to:
- determine, based on background noise at a near-end of a speaker, a noise estimate associated with speech emitted from the speaker;
- compare, by using a masking threshold below which presence of any speech is rendered inaudible, a speech intelligibility index, an articulation index, or a speech transmission index, the noise estimate to a speech level estimate for the speech emitted from the speaker;
- determine, based on the comparison, a gain value to be applied to the speaker to produce the speech at a specified level to maintain on-axis intelligibility with respect to the speaker;
- determine, based on applying a non-linear acoustic model, a modulated ultrasound level estimation of far-end speech;
- and
- apply the gain value and the modulated ultrasound level estimation to the speaker.
Dependencies are inferred from the claim text.