| 摘要 |
PROBLEM TO BE SOLVED: To enable high-accuracy identification of an operator with small data and arithmetic load.SOLUTION: A performer identification method includes the steps of: acquiring motion data about each given action of multiple performers; generating a feature vector for the motion data working as a sample, generating feature data with a lower-dimensional freedom degree m than the feature vector and storing the feature data in a storage part as sample data upon associating it with a performer; applying clustering the sample data, with respect to each performer, and obtaining cluster information that includes at least one of a feature point of a cluster in an m-dimensional space and its feature function; generating feature data with a lower dimension m than the feature vector generated after acquiring the motion data for the performer to be identified, and specifying the feature data as target data; and calculating a feature value of the target data on the basis of the target data in the m-dimensional space and at least one of the feature point of each cluster and its feature function therein, and identifying the performer on the basis of the feature value. |
