Information processing device and information processing method
View Patent ↗An information processing device includes: a processor configured to: obtain sensor data; determine partial data; input the sensor data to an input layer of a first neural network recognition model; input the partial data to an input layer of a second neural network recognition model; align second intermediate output data from a second intermediate layer of the second neural network recognition model, relative to first intermediate output data from a first intermediate layer of the first neural network recognition model; generate intermediate input data from the first intermediate output data and the second intermediate output data aligned; and input the intermediate input data to a third intermediate layer of the first neural network recognition model subsequent to the first intermediate layer.
1. An information processing device, comprising:
a processor configured to:
obtain sensor data from a sensor;
determine partial data which is a portion of the sensor data;
input the sensor data to an input layer of a first neural network recognition model;
input only the partial data of the sensor data to an input layer of a second neural network recognition model;
align second intermediate output data from a second intermediate layer of the second neural network recognition model, relative to first intermediate output data from a first intermediate layer of the first neural network recognition model;
generate intermediate input data from the first intermediate output data and the aligned second intermediate output data; and
input the intermediate input data to a third intermediate layer of the first neural network recognition model subsequent to the first intermediate layer, wherein
in aligning the second intermediate output data, a position or a size of an entire data array of the second intermediate output data is adjusted to a position or a size of a first region, the first region being a partial data array in a data array of the first intermediate output data.
2. The information processing device according to claim 1 , wherein
the sensor is an image sensor, and
the sensor data is an image.
3. The information processing device according to claim 1 , wherein
in aligning the second intermediate output data, the first region is determined according to a region of the partial data in the sensor data, and the position or the size of the entire data array of the second intermediate output data is adjusted to the position or the size of the determined first region.
4. The information processing device according to claim 1 , wherein
the processor is further configured to select the first intermediate layer from among a plurality of layers of the first neural network recognition model.
5. The information processing device according to claim 4 , wherein
the processor is further configured to select the first intermediate layer according to information which the first neural network recognition model and the second neural network recognition model have learned during training.
6. The information processing device according to claim 4 , wherein
the processor is further configured to select the first intermediate layer according to computational resource information on a computational resource of the information processing device.
7. The information processing device according to claim 1 , wherein
the processor is further configured to determine the partial data according to a parameter of the sensor.
8. The information processing device according to claim 1 , wherein
the processor is further configured to determine the partial data according to information indicating operation of or control over a movable object on which the sensor is mounted.
9. An information processing method, comprising:
obtaining sensor data from a sensor;
determining partial data which is a portion of the sensor data;
inputting the sensor data to an input layer of a first neural network recognition model;
inputting only the partial data of the sensor data to an input layer of a second neural network recognition model;
aligning second intermediate output data from a second intermediate layer of the second neural network recognition model, relative to first intermediate output data from a first intermediate layer of the first neural network recognition model;
generating intermediate input data from the first intermediate output data and the aligned second intermediate output data; and
inputting the intermediate input data to a third intermediate layer of the first neural network recognition model subsequent to the first intermediate layer, wherein
in aligning the second intermediate output data, a position or a size of an entire data array of the second intermediate output data is adjusted to a position or a size of a first region, the first region being a partial data array in a data array of the first intermediate output data.
10. A non-transitory computer-readable recording medium having a program recorded thereon for causing a computer to perform the information processing method according to claim 9 .