Nec Corporation

Sakiko Mishima

Reishi Kondo

Takahiro Toizumi

Tatsuya Komatsu

Yu Kiyokawa

Kazutoshi Sagi

Provided are a sound model generation device and the like that make it possible to more easily generate a sound model capable of distinguishing sound events using a plurality of features. A concatenating unit concatenates a plurality of features of sound signals that are learning data and generates a concatenated feature. A learning unit learns the generated concatenated feature for generating a sound model for distinguishing a sound event from a sound signal.

Sound model generation device, sound model generation method, and recording medium

Provided is a noise estimation device capable of appropriately estimating the amount of noise in an observation signal. The noise estimation device includes: frequency analysis processing means for receiving an input of an observation signal that includes a moving object sound output from an moving object and noise and transforming the observation signal into a feature in each of time-frequency domains; noise range estimation means for estimating a first feature in a first time-frequency domain to which only the noise belongs based on acoustic characteristic information of the moving object sound and the feature; and amount-of-noise estimation means for estimating an amount of noise in a second time-frequency domain to which the moving object sound belongs based on the first feature.

Noise estimation device, moving object sound detection device, noise estimation method, moving object sound detection method, and non-transitory computer-readable medium

The acoustic feature extraction meansextracts an acoustic feature, using predetermined parameters, from an acoustic pattern obtained as a result of processing on an acoustic signal. The language vector calculation meanscalculates a language vector from a given label that represents an attribute of a source of the acoustic signal and that is associated with the acoustic pattern. The similarity calculation meanscalculates a similarity between the acoustic feature and the language vector. The parameter update meanslearns parameters so that the similarity becomes larger, and updates the predetermined parameters to the parameters obtained by learning.

Learning device and pattern recognition device

An extraction unit extracts sound pressure information from a spectrogram. A binarization unit carries out binarization on the extracted sound pressure information in order to generate an event mask indicating a time period in which an audio event exists.

Growing community of inventors

Tokyo, Japan

Sakiko Mishima