ATHLETIC PERFORMANCE ESTIMATION APPARATUS, ATHLETIC PERFORMANCE ESTIMATION METHOD, AND PROGRAM
An athletic performance estimation apparatus evaluates athletic performance from the movement of eyes when observing the movement of a target. A control circuitry ( 11 ) selects a measurement condition. A video presentation circuitry ( 12 ) presents video according to the measurement condition to an observed person. An eyeball movement measurement circuitry ( 13 ) measures the movement of eyes of the observed person who is observing the movement of a target. An analysis circuitry ( 14 ) obtains a feature value based on an eyeball movement, from time series information on the movement of the eyes. A normalization circuitry ( 15 ) integrates the obtained feature value with a feature value obtained under a different measurement condition and performs normalization. An estimation circuitry ( 16 ) estimates the athletic performance of the observed person from a normalized feature value.
1 . An athletic performance estimation apparatus comprising:
an analysis circuitry that obtains a feature value based on an eyeball movement of an observed person who is observing movement of a target; and
an estimation circuitry that estimates athletic performance of the observed person from the feature value obtained from the observed person, based on a predetermined relationship between the feature value based on the eyeball movement and a level of athletic performance.
2 . The athletic performance estimation apparatus according to claim 1 , wherein
the feature value is time-series information including a time when a saccadic eye movement occurs and an amplitude thereof; and
the estimation circuitry estimates the athletic performance of the observed person so that a case in which the time and the amplitude included in the feature value and within a predetermined time period immediately before the target moves has a correlation corresponds to a higher level of athletic performance than otherwise.
3 . The athletic performance estimation apparatus according to claim 2 , further comprising:
a classifier that has been trained in advance to classify the feature value obtained from the observed person into either a first category or a second category, the classification being performed using both the feature value obtained from a person falling under the first category and the feature value obtained from a person falling under the second category, the first category being of high athletic performance, the second category being of lower athletic performance than the first category,
wherein the estimation circuitry obtains a classification result as an estimation result of the athletic performance of the observed person, the classification result being obtained by inputting the feature value obtained from the observed person into the classifier.
4 . The athletic performance estimation apparatus according to claim 2 , wherein
the feature value is a normalized feature value obtained by integrating a first feature value with a second feature value, the first feature value being obtained from the observed person when the target moves in a first direction, the second feature value being obtained from the observed person when the target moves in a second direction different from the first direction.
5 . An athletic performance estimation method, comprising:
an analysis circuitry obtaining a feature value based on eyeball movement of an observed person who is observing movement of a target; and
an estimation circuitry estimating athletic performance of the observed person from the feature value obtained from the observed person, based on a predetermined relationship between the feature value based on the eyeball movement and a level of athletic performance.
6 . A non-transitory computer-readable recording medium which stores a program for causing a computer to function as the athletic performance estimation apparatus according to claim 1 .