Information identification processing apparatus, information identification processing method, image identification system, and storage medium
There is provided with an information processing apparatus. An extracting unit extracts a feature vector of a subject in an image. A generating unit configured to generate, based on the feature vector, a post-conversion vector including a converted vector resulting from conversion of the feature vector using a random number, the post-conversion vector being a vector whose number of elements is determined according to the random number and is larger than a number of elements of the feature vector.
1. An information processing apparatus comprising:
at least one processor; and
at least one memory having stored thereon instructions which, when executed by the at least one processor, cause the information processing apparatus at least to:
extract at least one feature vector corresponding to a subject in an image;
generate a different seed value for each feature vector of the at least one feature vector;
generate a different random sequence for each of the seed values; and
calculate at least one post-conversion vector by inserting each seed value into a predetermined position of a corresponding corrected vector,
wherein each corrected vector is obtained by performing an operation on elements positionally corresponding to each other between a corresponding feature vector and a corresponding random sequence, and
wherein the operation enables reconstruction of each feature vector of the at least one feature vector using a corresponding seed value and a corresponding post-conversion vector.
2. The information processing apparatus according to claim 1 , wherein the instructions, when executed by the at least one processor, further cause the information processing apparatus at least to:
calculate a converted vector in which a value of each of the elements is determined according to the random sequence,
wherein the corrected vector is generated by updating a value of each of the elements of one of the at least one feature vector and the converted vector that has a larger number of elements, or a value of each of the elements of one of the at least one feature vector and the converted vector if numbers of elements of the at least one feature vector and the converted vector are equal, based on a value of a positionally corresponding element of the other of the at least one feature vector and the converted vector.
3. The information processing apparatus according to claim 2 ,
wherein the seed value is calculated based on the at least one feature vector.
4. An image identification system comprising:
the information processing apparatus according to claim 1 ;
at least one second processor; and
at least one second memory storing instructions which, when executed by the at least one second processor, causes the image identification system at least to:
perform class determination based on at least one feature vector, wherein the performing the class determination includes:
recalculating the at least one feature vector based on the post-conversion vector, and
perform class determination for the subject, using the recalculated at least one feature vector.
5. The image identification system according to claim 4 , wherein the instructions, when executed by the at least one second processor, further cause the image identification system at least to:
generate, based on the at least one feature vector, a seed value used when generating the random sequence,
wherein the performing the class determination further includes:
obtaining a seed value of the random sequence based on the post-conversion vector,
wherein the class determination for the subject is performed if the obtained seed value and the generated seed value based on the recalculated at least one feature vector are identical.
6. The information processing apparatus according to claim 1 ,
wherein a number of elements of each post-conversion vector is determined according to the corresponding random sequence and is larger than a number of elements of the corresponding feature vector.
7. The information processing apparatus according to claim 1 , wherein each random sequence has a different column length.
8. An information processing method comprising:
extracting at least one feature vector corresponding to a subject in an image;
generating a different seed value for each feature vector of the at least one feature vector;
generating a different random sequence for each of the seed values; and
calculating at least one post-conversion vector by inserting each seed value into a predetermined position of a respective corrected vector,
wherein each corrected vector is obtained by performing an operation on elements positionally corresponding to each other between a corresponding feature vector and a corresponding random sequence, and
wherein the operation enables reconstruction of each feature vector of the at least one feature vector using a corresponding seed value and a corresponding post-conversion vector.
9. A non-transitory computer readable storage medium storing a program that, when executed by a computer, causes the computer to perform an information processing method comprising:
extracting at least one feature vector corresponding to a subject in an image;
generating a different seed value for each feature vector of the at least one feature vector;
generating a different random sequence for each of the seed values; and
calculating at least one post-conversion vector by inserting each seed value into a predetermined position of a respective corrected vector,
wherein each corrected vector is obtained by performing an operation on elements positionally corresponding to each other between a corresponding feature vector and a corresponding random sequence, and
wherein the operation enables reconstruction of each feature vector of the at least one feature vector using a corresponding seed value and a corresponding post-conversion vector.