Information processing apparatus, method, and program
AssigneeFUJIFILM Corporation
InventorsHirokazu KAMEYAMA, Kouji YAMAGUCHI
Life of this patent
Estimating the term…
Abstract
Disclosed is a method for accurately and efficiently detecting a modality of input data, including the steps of projecting the input data into a plurality of projection data using each of a plurality of transformation matrix groups U1·(Σ12U2T), generating a plurality of inverse projection data by performing inverse projection of the transformation matrix groups on the plurality of generated projection data, calculating a correlation between the input data and the generated inverse projection data with respect to each transformation matrix group U1·(Σ12U2T), and identifying a modality represented by a transformation matrix group having a highest calculated correlation as the modality of the input data.
Claims12
An information processing apparatus for identifying a modality of input data using a plurality of transformation matrix groups, each including a first projection matrix for projecting the input data into a space vector and a second projection matrix for projecting the space vector into projection data having a smaller dimension number than that of the space vector, learned through a plurality of sample data, each having the same modality for a non-identification target and a different modality for an identification target, with respect to each of the different modalities, the apparatus comprising:
- a first projection means for projecting the input data into a plurality of space vectors using a plurality of the first projection matrices;
- a second projection means for projecting the input data into a plurality of projection data using the plurality of transformation matrix groups and generating inverse space vectors by performing inverse projection of the second transformation matrix on the plurality of projection data;
- a correlation calculation means for calculating a correlation between the space vector generated by the first projection means and the inverse space vector generated by the second projection means with respect to each transformation matrix group;
- and
- an identification means for identifying a modality represented by a transformation matrix group having a highest correlation calculated by the correlation calculation means as the modality of the input data.
An information processing apparatus for identifying a modality of input data using a plurality of transformation matrix groups, each including a first projection matrix for projecting the input data into a space vector and a second projection matrix for projecting the space vector into projection data having a smaller dimension number than that of the space vector, learned through a plurality of sample data, each having the same modality for a non-identification target and a different modality for an identification target, with respect to each of the different modalities, the apparatus comprising:
- a first projection means for projecting the input data into a plurality of projection data using the plurality of transformation matrix groups;
- a second projection means for generating a plurality of inverse projection data by performing inverse projection of the transformation matrix groups on the plurality of projection data generated by the first projection means;
- a correlation calculation means for calculating a correlation between the input data and the inverse projection data generated by the second projection means with respect to each transformation matrix group;
- and
- an identification means for identifying a modality represented by a transformation matrix group having a highest correlation calculated by the correlation calculation means as the modality of the input data.
An information processing method for identifying a modality of input data using a plurality of transformation matrix groups, each including a first projection matrix for projecting the input data into a space vector and a second projection matrix for projecting the space vector into projection data having a smaller dimension number than that of the space vector, learned through a plurality of sample data, each having the same modality for a non-identification target and a different modality for an identification target, with respect to each of the different modalities, the method comprising the steps of:
- projecting the input data into a plurality of space vectors using a plurality of the first projection matrices;
- projecting the input data into a plurality of projection data using the plurality of transformation matrix groups and generating inverse space vectors by performing inverse projection of the second transformation matrix on the plurality of projection data;
- calculating a correlation between the generated space vector and the generated inverse space vector with respect to each transformation matrix group;
- and
- identifying a modality represented by a transformation matrix group having a highest calculated correlation as the modality of the input data.
An information processing method for identifying a modality of input data using a plurality of transformation matrix groups, each including a first projection matrix for projecting the input data into a space vector and a second projection matrix for projecting the space vector into projection data having a smaller dimension number than that of the space vector, learned through a plurality of sample data, each having the same modality for a non-identification target and a different modality for an identification target, with respect to each of the different modalities, the method comprising the steps of:
- projecting the input data into a plurality of projection data using the plurality of transformation matrix groups;
- generating a plurality of inverse projection data by performing inverse projection of the transformation matrix groups on the plurality of generated projection data;
- calculating a correlation between the input data and the generated inverse projection data with respect to each transformation matrix group;
- and
- identifying a modality represented by a transformation matrix group having a highest calculated correlation as the modality of the input data.
A non-transitory computer readable recording medium on which is recorded an information processing program for causing a computer to identify a modality of input data using a plurality of transformation matrix groups, each including a first projection matrix for projecting the input data into a space vector and a second projection matrix for projecting the space vector into projection data having a smaller dimension number than that of the space vector, learned through a plurality of sample data, each having the same modality for a non-identification target and a different modality for an identification target, with respect to each of the different modalities, the program causing the computer to perform the steps of:
- projecting the input data into a plurality of space vectors using a plurality of the first projection matrices;
- projecting the input data into a plurality of projection data using the plurality of transformation matrix groups and generating inverse space vectors by performing inverse projection of the second transformation matrix on the plurality of projection data;
- calculating a correlation between the generated space vector and the generated inverse space vector with respect to each transformation matrix group;
- and
- identifying a modality represented by a transformation matrix group having a highest calculated correlation as the modality of the input data.
A non-transitory computer readable recording medium on which is recorded an information processing program for causing a computer to identify a modality of input data using a plurality of transformation matrix groups, each including a first projection matrix for projecting the input data into a space vector and a second projection matrix for projecting the space vector into projection data having a smaller dimension number than that of the space vector, learned through a plurality of sample data, each having the same modality for a non-identification target and a different modality for an identification target, with respect to each of the different modalities, the program causing the computer to perform the steps of:
- projecting the input data into a plurality of projection data using the plurality of transformation matrix groups;
- generating a plurality of inverse projection data by performing inverse projection of the transformation matrix groups on the plurality of generated projection data;
- calculating a correlation between the input data and the generated inverse projection data with respect to each transformation matrix group;
- and
- identifying a modality represented by a transformation matrix group having a highest calculated correlation as the modality of the input data.
Dependencies are inferred from the claim text.