US 12,085,391 B2
Positioning apparatus, positioning method, and program
Takahiro Tsujii, Kanagawa (JP)
Assigned to SONY CORPORATION, Tokyo (JP)
Appl. No. 16/972,659
Filed by SONY CORPORATION, Tokyo (JP)
PCT Filed Jun. 19, 2019, PCT No. PCT/JP2019/024214
§ 371(c)(1), (2) Date Dec. 7, 2020,
PCT Pub. No. WO2020/008878, PCT Pub. Date Jan. 9, 2020.
Claims priority of provisional application 62/693,019, filed on Jul. 2, 2018.
Prior Publication US 2021/0247189 A1, Aug. 12, 2021
Int. Cl. G01C 21/16 (2006.01); G06N 20/00 (2019.01)
CPC G01C 21/188 (2020.08) [G06N 20/00 (2019.01)] 17 Claims
OG exemplary drawing
 
1. A positioning apparatus, comprising:
a movement vector estimator configured to estimate a movement vector using a machine learning model,
wherein the estimated movement vector indicates a movement amount and a movement direction of a device in a predetermined time period,
wherein the movement vector is estimated on a basis of input data to the machine learning model, the input data including acceleration of the device in the predetermined time period and an angular velocity of the device in the predetermined time period,
wherein the acceleration of the device is detected by an acceleration sensor, and
wherein the angular velocity of the device is detected by an angular velocity sensor;
an integration section configured to
integrate the estimated movement vector, and
calculate a relative position of the device for the predetermined time period with respect to a specified reference position in a real space; and
a display control unit configured to initiate display of a trajectory of the relative position of the device for the predetermined time period with respect to the specified reference position in the real space,
wherein the machine learning model is trained in advance on a basis of acceleration of one or more devices in each of a plurality of prior predetermined time periods and an angular velocity of the one or more devices in each of the plurality of prior predetermined time periods, and
wherein the movement vector estimator and the integration section are each implemented via at least one processor.