Abstract: The purpose of the present invention is to enable learning of a neural network for extracting features of images having high robustness from an undiscriminating image region while minimizing the number of parameters of a pooling layer. A parameter learning unit 130 learns parameters of each layer in a convolutional neural network configured by including a fully convolutional layer for performing convolution of an input image to output a feature tensor of the input image, a weighting matrix estimation layer for estimating a weighting matrix indicating a weighting of each element of the feature tensor, and a pooling layer for extracting a feature vector of the input image based on the feature tensor and the weighting matrix.

Abstract: The present invention enables detection of a sponsorship credit display segment in a broadcast program with higher precision.

Abstract: The present invention enables detection of a sponsorship credit display in a broadcast program with higher precision. A sponsorship credit display detection device 100 according to the present invention includes: a CM segment detection unit 120 that extracts a cut point, which is a time point where a frame in which the volume of an audio signal of a broadcast program is less than a volume threshold value and the amount of change from a previous frame is at least a pixel change threshold value is played, and detects a CM segment by comparing an interval of the extracted cut point with a CM defined length; a sponsorship credit display segment estimation unit 130 that estimates, as a sponsorship credit display segment, a predetermined time period before or after at least one continuous CM segment detected by the CM segment detection unit 120; and an output unit 140 that outputs information indicating the sponsorship credit display segment.

Abstract: Low-dimensional feature values with which semantic factors of content are ascertained are generated from relevance between sets of two types of content. Based on a relation indicator indicating a pair of groups indicating which groups are related to first types of content groups among second types of content groups, an initial feature value extracting unit 11 extracts initial feature values of the first type of content and the second type of content. A content pair selecting unit 12 selects a content pair by selecting one first type of content and one second type of content from each pair of groups indicated by the relation indicator. A feature value conversion function generating unit 13 generates feature conversion functions 31 of converting the initial feature values into low-dimensional feature values based on the content pair selected from each pair of groups.

Abstract: A signal retrieval device includes a modification unit and a signal retrieval unit. The modification unit modifies a value of an attribute of a target represented by an input signal or a stored signals stored in a signal storage unit or a value of an attribute relating to a signal generation source of the input signal to acquire a plurality of modified values of the attribute. The signal retrieval unit retrieves a stored signal of the stored signals similar to the input signal using the input signal or the stored signals in which the attribute is modified according to each of the plurality of modified values of the attribute acquired by the modification unit.

Abstract: Clustering can be performed using a self-expression matrix in which noise is suppressed. A self-expression matrix is calculated that minimizes an objective function that is for obtaining, from among matrices included in a predetermined matrix set, a self-expression matrix whose elements are linear weights when data points in a data set are expressed by linear combinations of points, the objective function being represented by a term for obtaining the residual between data points in the data set and data points expressed by linear combinations of points using the self-expression matrix, a first regularization term that is multiplied by a predetermined weight and is for reducing linear weights of the data points that have a large Euclidean norm in the self-expression matrix, and a second regularization term for the self-expression matrix. A similarity matrix defined by the calculated self-expression matrix is then calculated.

Abstract: Information related to CMs included in a broadcast program can be automatically added. A CM information generation device 100 includes: a CM section detection unit 120 that detects one or more CM sections within a broadcast program by comparing the volume of the broadcast program with a volume threshold; a CM detection list generation unit 150 that generates a CM detection list describing company names of companies that have advertised detected CMs, which are CMs in the CM sections detected by the CM section detection unit 120, by cross-referencing the detected CMs with CM masters that have been associated with company names of advertisers in advance; a company name list generation unit 170 that generates a company name list describing company names that are specified by a sponsorship credit display indicating sponsors of the broadcast program; and a CM information generation unit 180 that generates CM information related to the detected CMs by comparing the CM detection list with the company name list.

Abstract: The chroma of a specific color can be emphasized with respect to an object without affecting the color of illumination light. An illumination light generator generates illumination light to be radiated to the object by adding or subtracting, to or from a reference illumination light spectrum that is an illumination spectrum serving as a reference, an element spectrum in accordance with designated conditions with respect to chroma adjustment from among a plurality of element spectra that are spectra for being added or subtracted to or from the reference illumination light spectrum and for performing chroma emphasis of a specific color without affecting an illumination light color.

Abstract: A CM section within a broadcast program is detected with high accuracy. A CM section detection device 100 includes: a CM section detection unit 120 that detects one or more CM sections by comparing a volume of a broadcast program with a volume threshold, and generates detected CM sections representing the CM sections that have been detected; and a CM section correction unit 140 that corrects the detected CM sections based on a sponsorship credit display section that is a section which is included in the broadcast program and in which a sponsorship credit indicating a sponsor of the broadcast program is displayed.

Abstract: A signal output unit outputs an acquired signal. A signal accumulation unit accumulates signals. A signal attribute value display unit displays a value of an attribute related to an element constituting a target represented by the acquired signal or a signal generation source in a state in which an instruction for changing the value of the attribute is able to be received. A changed attribute value acquisition unit acquires a changed value of the attribute when the instruction for changing the value of the attribute is received. A signal retrieval unit retrieves, from the signals accumulated in the signal accumulation unit, a signal similar to a changed signal when the value of the attribute has been changed on the basis of the changed value of the attribute acquired by the changed attribute value acquisition unit.

Abstract: A signal output unit outputs an acquired signal. A signal attribute value display unit displays a value of an attribute related to an element constituting a target represented by the acquired signal or a signal generation source in a state in which a change instruction of the value of the attribute is able to be received. A changed attribute value acquisition unit acquires a changed value of the attribute when the change instruction of the value of the attribute is received. A change unit changes the value of the attribute for which the change instruction has been received on the basis of the changed value of the attribute acquired by the changed attribute value acquisition unit. A changed signal output unit outputs a changed signal in which the value of the attribute has been changed.

Abstract: An identifier of query content is accurately determined. A content determining device 100 includes an input unit 2 that inputs query content, a storage unit 1 that stores a plurality of pieces of master content, and a content determining unit 4 that determines a region where feature values of two pieces of master content out of the plurality of pieces of master content do not match each other, calculates a matching feature count which is a count of feature values of the region that match feature values in a corresponding region of the query content, for each of the two pieces of master content, and determines an identifier of the query content on the basis of the matching feature count of each of the pieces of master content.

Abstract: Clustering can be performed using a self-expression matrix in which noise is suppressed. A self-expression matrix is calculated that minimizes an objective function that is for obtaining, from among matrices included in a predetermined matrix set, a self-expression matrix whose elements are linear weights when data points in a data set are expressed by linear combinations of points, the objective function being represented by a term for obtaining the residual between data points in the data set and data points expressed by linear combinations of points using the self-expression matrix, a first regularization term that is multiplied by a predetermined weight and is for reducing linear weights of the data points that have a large Euclidean norm in the self-expression matrix, and a second regularization term for the self-expression matrix. A similarity matrix defined by the calculated self-expression matrix is then calculated.

Abstract: Low-dimensional feature values with which semantic factors of content are ascertained are generated from relevance between sets of two types of content. Based on a relation indicator indicating a pair of groups indicating which groups are related to first types of content groups among second types of content groups, an initial feature value extracting unit 11 extracts initial feature values of the first type of content and the second type of content. A content pair selecting unit 12 selects a content pair by selecting one first type of content and one second type of content from each pair of groups indicated by the relation indicator. A feature value conversion function generating unit 13 generates feature conversion functions 31 of converting the initial feature values into low-dimensional feature values based on the content pair selected from each pair of groups.

Abstract: The purpose of the present invention is to enable learning of a neural network for extracting features of images having high robustness from an undiscriminating image region while minimizing the number of parameters of a pooling layer. A parameter learning unit 130 learns parameters of each layer in a convolutional neural network configured by including a fully convolutional layer for performing convolution of an input image to output a feature tensor of the input image, a weighting matrix estimation layer for estimating a weighting matrix indicating a weighting of each element of the feature tensor, and a pooling layer for extracting a feature vector of the input image based on the feature tensor and the weighting matrix.

Abstract: The present invention enables the structure of a neural network to be quantitatively analyzed. An analyzing unit calculates, for each of combinations of a dimension of input data and a cluster, a sum of squared errors between an output of each unit belonging to the cluster when a value of the dimension of the input data is replaced with an average value of the dimension of the input data included in learning data and an output of each unit belonging to the cluster for the input data before replacement as a relationship between the combinations, and calculates, for each of combinations of the cluster and a dimension of output data, a squared error between the value of the dimension of the output data when an output value of each unit belonging to the cluster is replaced with an average output value of each unit of the cluster when the input data included in the learning data was input and the value of the dimension of the output data before replacement as a relationship between the combinations.

Abstract: The present invention enables detection of a sponsorship credit display in a broadcast program with higher precision. A sponsorship credit display detection device 100 according to the present invention includes: a CM segment detection unit 120 that extracts a cut point, which is a time point where a frame in which the volume of an audio signal of a broadcast program is less than a volume threshold value and the amount of change from a previous frame is at least a pixel change threshold value is played, and detects a CM segment by comparing an interval of the extracted cut point with a CM defined length; a sponsorship credit display segment estimation unit 130 that estimates, as a sponsorship credit display segment, a predetermined time period before or after at least one continuous CM segment detected by the CM segment detection unit 120; and an output unit 140 that outputs information indicating the sponsorship credit display segment.

Abstract: The present invention enables detection of a sponsorship credit display segment in a broadcast program with higher precision.

Abstract: It is possible to analyze what type of input/output mapping is mainly performed in the entire hidden layer of a neural network. A relationship analysis unit 30 calculates strength of a relationship of each combination of a dimension of the input data and a unit of the neural network and calculates strength of a relationship of each combination of the unit and a dimension of the output data. A role analysis unit 32 calculates a relationship between a prescribed number of types of roles and the unit and a relationship between the prescribed number of types of roles, the dimension of the input data, and the dimension of the output data on the basis of the strength of the relationship of each combination of the dimension of the input data and the unit of the neural network and the strength of the relationship of each combination of the unit and the dimension of the output data.

Abstract: The present invention enables detection of a sponsorship credit display in a broadcast program with higher precision. A sponsorship credit display detection device 100 according to the present invention includes: a learning data creation unit 140 that creates, as learning data, from a broadcast program in which the sponsorship credit display has been detected, a still image where the sponsorship credit display is displayed and a still image where the sponsorship credit display is not displayed; a learning unit 150 that uses the learning data created by the learning data creation unit 140 to learn a parameter to be applied to a detection model for detecting a sponsorship credit display in a broadcast program; and a sponsorship credit display detection unit 170 that uses the detection model to which the parameter learned by the learning unit 150 has been applied to detect a sponsorship credit display in a recognition target broadcast program.