Information processing apparatus, information processing method, non-transitory computer readable medium, and learning model

An information processing method performed by at least one processor and comprising:

• acquiring map information from an external source;
• acquiring prestored movement trajectories from a storage of an information processing apparatus;
• assigning soft labels corresponding to the map information to the prestored movement trajectories;
• acquiring prestored movement information from the storage;
• outputting labels based on the soft labels and the prestored movement information;
• generating a plurality of data sets, each data set of the plurality of data sets including a label among the labels associated with a corresponding prestored movement trajectory among the prestored movement trajectories and corresponding prestored movement information;
• acquiring a movement trajectory of a user;
• deriving, from the movement trajectory of the user, movement information indicating features relating to movement;
• and
• estimating, using a learning model, a mode of transport of the user from the movement trajectory and the movement information,
• wherein the learning model is trained based on the plurality of data sets,
• wherein the learning model includes a first network, which is composed of a first branch and a second branch, and a second network that follows the first network,
• the first branch generates feature amounts of the movement trajectory from the movement trajectory,
• the second branch generates feature amounts of the movement information from the movement information, and
• the second network is configured to generate combined feature amounts by combining the feature amounts of the movement trajectory and the feature amounts of the movement information, and output data indicating the mode of transport of the user from the combined feature amounts.

A non-transitory computer readable medium storing a computer program for causing a computer to execute processing comprising:

• a training process for:
• acquiring map information from an external source,
• acquiring prestored movement trajectories from a storage of an information processing apparatus,
• assigning soft labels corresponding to the map information to the prestored movement trajectories,
• acquiring prestored movement information from the storage,
• outputting labels based on the soft labels and the prestored movement information, and
• generating a plurality of data sets, each data set of the plurality of data sets including a label among the labels associated with a corresponding prestored movement trajectory among the prestored movement trajectories and corresponding prestored movement information;
• an acquisition process for acquiring a movement trajectory of a user;
• a derivation process for deriving, from the movement trajectory of the user, movement information indicating features relating to movement;
• and
• an estimating process for estimating, using a learning model, a mode of transport of the user from the movement trajectory and the movement information,
• wherein the learning model is trained based on the plurality of data sets,
• wherein the learning model includes a first network, which is composed of a first branch and a second branch, and a second network that follows the first network,
• the first branch generates feature amounts of the movement trajectory from the movement trajectory,
• the second branch generates feature amounts of the movement information from the movement information, and
• the second network is configured to generate combined feature amounts by combining the feature amounts of the movement trajectory and the feature amounts of the movement information, and output data indicating the mode of transport of the user from the combined feature amounts.

An information processing apparatus comprising:

• at least one memory configured to store program code;
• and
• at least one processor configured to operate as instructed by the program code, the program code including:
• training code configured to cause at least one of the at least one processor to:
• acquire map information from an external source,
• acquire prestored movement trajectories from a storage of the information processing apparatus,
• assign soft labels corresponding to the map information to the prestored movement trajectories,
• acquire prestored movement information from the storage,
• output labels based on the soft labels and the prestored movement information,
• generate a plurality of data sets, each data set of the plurality of data sets including a label among the labels associated with a corresponding prestored movement trajectory among the prestored movement trajectories and corresponding prestored movement information;
• estimating code configured to cause at least one of the at least one processor to estimate, using a learning model, a mode of transport of a user from the movement trajectory and the movement information,
• wherein the learning model is trained based on the plurality of data sets,
• wherein the learning model comprises:
• a first network including:
• a first branch configured to:
• receive a movement trajectory of the user, and
• generate feature amounts of the movement trajectory, and
• a second branch configured to:
• receive movement information relating to movement derived from the movement trajectory, and
• generate feature amounts of the movement information;
• and
• a second network configured to generate combined feature amounts by combining the feature amounts of the movement trajectory and the feature amounts of the movement information and to output data indicating a mode of transport of the user from the combined feature amounts.