Psd optimization apparatus, psd optimization method, and program

A power spectral density (PSD) optimization method including a PSD updating step, in which,

• with uS as a variable representing a target sound PSD, uIN as a variable representing an interference noise PSD, and uBN as a variable representing a background noise PSD,
• a PSD optimization device takes a target sound PSD input value {circumflex over ( )}φS(ω, τ), an interference noise PSD input value {circumflex over ( )}φIN(ω, τ), and a background noise PSD input value {circumflex over ( )}φBN(ω, τ) as input, and generates a target sound PSD output value φS(ω, τ), an interference noise PSD output value {circumflex over ( )}φIN(ω, τ), and a background noise PSD output value {circumflex over ( )}φBN(ω, τ), by solving an optimization problem for a cost function relating to the variable uS, the variable uIN, and the variable uBN,
• wherein the optimization problem for the cost function is defined using at least one of:
• a constraint based on difference in sound source enhancement characteristics of a beamformer in accordance with a degree of inclusion of a target sound, with regard to a PSD of signals beamformed in a target sound direction-of-arrival,
• a convex cost term based on the difference in sound source enhancement characteristics of the beamformer in accordance with the degree of inclusion of the target sound, with regard to the PSD of signals beamformed in the target sound direction-of-arrival,
• a constraint based on the difference in sound source enhancement characteristics of the beamformer in a frequency direction, with regard to the PSD of signals beamformed in the target sound direction-of-arrival, and
• a convex cost term based on the difference in sound source enhancement characteristics of the beamformer in the frequency direction, with regard to the PSD of signals beamformed in the target sound direction-of-arrival.

A non-transitory computer-readable medium having computer-readable instructions stored thereon, which, when executed, cause a computer including a memory and a processor to execute a set of operations, comprising:

• obtaining a target sound PSD input value {circumflex over ( )}φS(ω, τ), an interference noise PSD input value {circumflex over ( )}φIN(ω, τ), and a background noise PSD input value {circumflex over ( )}φBN(ω, τ) as input, with uS as a variable representing a target sound PSD, uIN as a variable representing an interference noise PSD, and uBN as a variable representing a background noise PSD, and
• generating a target sound PSD output value φS(ω, τ), an interference noise PSD output value φIN(ω, τ), and a background noise PSD output value φBN(ω, τ), by solving an optimization problem for a cost function relating to the variable uS, the variable uIN, and the variable uBN,
• wherein the optimization problem for the cost function is defined using at least one of:
• a constraint based on difference in sound source enhancement characteristics of a beamformer in accordance with a degree of inclusion of a target sound, with regard to a PSD of signals beamformed in a target sound direction-of-arrival,
• a convex cost term based on the difference in sound source enhancement characteristics of the beamformer in accordance with the degree of inclusion of the target sound, with regard to the PSD of signals beamformed in the target sound direction-of-arrival,
• a constraint based on the difference in sound source enhancement characteristics of the beamformer in a frequency direction, with regard to the PSD of signals beamformed in the target sound direction-of-arrival, and
• a convex cost term based on the difference in sound source enhancement characteristics of the beamformer in the frequency direction, with regard to the PSD of signals beamformed in the target sound direction-of-arrival.