Estimation device, estimation system, estimation method, and program recording medium
An estimation device that includes an extraction unit that acquires sensor data from a sensor installed on footwear and extracts a gait feature quantity characteristic of walking in the footwear by using the sensor data, and an estimation unit that estimates a type of the footwear based on the gait feature quantity extracted by the extraction unit.
1. An estimation device comprising:
at least one memory storing instructions; and
at least one processor connected to the at least one memory and configured to execute the instructions to:
acquire sensor data including acceleration and angular velocity from a sensor installed on a footwear worn by a user;
extract a gait feature quantity characteristic of walking in the footwear by using time-series data of the sensor data including the acceleration and the angular velocity;
estimate a heel height of the footwear, the heel height being a relative height of a heel of the footwear with respect to a footrest portion of a sole of the footwear, based on the gait feature quantity characteristic of walking in the footwear by using a second model obtained by machine learning the gait feature quantity characteristic of walking in the footwear with the heel height of the footwear used as a second label;
estimate a type of the footwear, the type of footwear being exercise shoes or high heels, based on the gait feature quantity relevant to a heel height by using a first model obtained by machine learning the gait feature quantity characteristic of walking in the footwear with the heel height of the footwear used as a label of exercise shoes or high heels; and
output data of a video moving image and advice regarding a way of walking and a posture relevant to an ideal gait with the type of the footwear so as to make a decision to improve the walking with the footwear.
2. The estimation device according to claim 1 , wherein
the at least one processor is configured to execute the instructions to
generate time-series data of a gait parameter by using the sensor data,
generate gait waveform data by normalizing the time-series data of the gait parameter to a gait cycle, and
extract the gait feature quantity characteristic of walking in the footwear from the gait waveform data.
3. The estimation device according to claim 2 , wherein
the at least one processor is configured to execute the instructions to
generate time-series data of a spatial angle by using the sensor data,
detect a middle time point of a stance phase from the time-series data of the spatial angle,
extract waveform data from a time zone between a middle time point of a preceding stance phase and a middle time point of a subsequent stance phase for continuous stance phases in the time-series data of the gait parameter, and
generate the gait waveform data by normalizing the extracted waveform data to the gait cycle.
4. The estimation device according to claim 3 , wherein
the at least one processor is configured to execute the instructions to extract at least one gait feature quantity of
a characteristic gait feature quantity in an initial swing period-extracted from gait waveform data obtained from time-series data of acceleration in a traveling direction of a pedestrian walking in the footwear,
a characteristic gait feature quantity in a terminal swing period-extracted from gait waveform data obtained from time-series data of acceleration in a gravity direction,
a characteristic gait feature quantity in at least one of heel separation, kicking, and the terminal swing period-extracted from gait waveform data obtained from time-series data of an angular velocity around a lateral axis of the pedestrian walking in the footwear,
a characteristic gait feature quantity in a period from a mid-stance period to the terminal swing period extracted from gait waveform data obtained from time-series data of an angle around the lateral axis of the pedestrian, and
a characteristic gait feature quantity in at least one of the terminal swing period and a heel rocker extracted from gait waveform data obtained from time-series data of a velocity in the gravity direction.
5. The estimation device according to claim 1 , wherein
the at least one processor is configured to execute the instructions to estimate a heel height of the footwear based on the gait feature quantity by using a second model obtained by machine learning the gait feature quantity characteristic of walking in the footwear with the heel height of the footwear used as a label.
6. An estimation system comprising:
the estimation device according to claim 1 ; and
a data acquisition device that is installed on the footwear, measures a spatial acceleration and a spatial angular velocity, generates the sensor data based on the measured spatial acceleration and spatial angular velocity, and transmits the generated sensor data to the estimation device.
7. An estimation method performed by a computer, the method comprising:
acquiring sensor data including acceleration and angular velocity from a sensor installed on a footwear worn by a user;
extracting a gait feature quantity characteristic of walking in the footwear by using time-series data of the sensor data including the acceleration and the angular velocity;
estimating a heel height of the footwear, the heel height being a relative height of a heel of the footwear with respect to a footrest portion of a sole of the footwear, based on the gait feature quantity characteristic of walking in the footwear by using a second model obtained by machine learning the gait feature quantity characteristic of walking in the footwear with the heel height of the footwear used as a second label;
estimating a type of the footwear, the type of footwear being exercise shoes or high heels, based on the extracted gait feature quantity relevant to a heel height by using a first model obtained by machine learning the gait feature quantity characteristic of walking in the footwear with the heel height of the footwear used as a label of exercise shoes or high heels; and
outputting data of a video moving image and advice regarding a way of walking and a posture relevant an ideal gait with to the type of the footwear so as to make a decision to improve the walking with the footwear.
8. A non-transitory program recording medium having stored therein a program causing a computer to execute a process comprising:
acquiring sensor data including acceleration and angular velocity from a sensor installed on a footwear worn by a user;
extracting a gait feature quantity characteristic of walking in the footwear by using time-series data of the sensor data including the acceleration and the angular velocity;
estimating a heel height of the footwear, the heel height being a relative height of a heel of the footwear with respect to a footrest portion of a sole of the footwear, based on the gait feature quantity characteristic of walking in the footwear by using a second model obtained by machine learning the gait feature quantity characteristic of walking in the footwear with the heel height of the footwear used as a second label;
estimating a type of the footwear, the type of footwear being exercise shoes or high heels, based on the extracted gait feature quantity relevant to a heel height by using a first model obtained by machine learning the gait feature quantity characteristic of walking in the footwear with the heel height of the footwear used as a label of exercise shoes or high heels; and
outputting data of a video moving image and advice regarding a way of walking and a posture relevant an ideal gait with to the type of the footwear so as to make a decision to improve the walking with the footwear.
9. The estimation device according to claim 1 , wherein
the at least one processor is configured to execute the instructions to predict a change in the heel height over time.