Solid state image capturing system, solid state image capturing device, information processing device, image processing method, information processing method

1. A solid-state image capturing system, comprising:

a solid-state image capturing device; and

an information processing device, wherein

the solid-state image capturing device includes:

a first Deep-Neural-Network (DNN) processing unit that is configured to execute, on image data, a part of a DNN algorithm by a first DNN to generate a first result, and

the information processing device includes:

a second DNN processing unit that is configured to execute, on the first result acquired from the solid-state image capturing device, a remaining part of the DNN algorithm by a second DNN to generate a second result,

wherein the information processing device is configured to:

generate second control information which indicates completion of the execution of the second DNN, and

transmit the generated second control information to the solid-state image capturing device.

2. The solid-state image capturing system according to claim 1 , wherein

the first result includes a feature map that is output from an intermediate layer of the DNN algorithm.

3. The solid-state image capturing system according to claim 1 , wherein

the solid-state image capturing device further includes:

a first storage that is configured to store therein at least the part of the DNN algorithm for execution of the first DNN, and

the information processing device further includes:

a second storage that is configured to store therein at least the remaining part of the DNN algorithm for the execution of the second DNN.

4. The solid-state image capturing system according to claim 3 , wherein

the DNN algorithm to be executed on the image data is stored in the first storage and the second storage.

5. The solid-state image capturing system according to claim 1 , wherein

the solid-state image capturing device further includes:

a first control unit that is configured to control the first DNN processing unit, and

the information processing device further includes:

a second control unit that is configured to control the second DNN processing unit.

6. The solid-state image capturing system according to claim 5 , wherein

the first control unit configured to generate first control information including information on the first DNN, and transmit the generated first control information to the second control unit, and

the second control unit configured to generate the second control information including information on the second DNN, and transmit the generated second control information to the first control unit.

7. The solid-state image capturing system according to claim 6 , wherein

the first control unit configured to control the first DNN processing unit based on the second control information, and

the second control unit configured to control the second DNN processing unit based on the first control information.

8. The solid-state image capturing system according to claim 5 , wherein

the first control unit configured to transmit, to the second control unit, an execution completion notification of the first DNN processing unit, and

the second control unit configured to transmit, to the first control unit, an execution completion notification of the second DNN processing unit.

9. The solid-state image capturing system according to claim 1 , wherein

the information processing device includes one of an application processor and a cloud server.

10. A solid-state image capturing device, comprising:

a Deep-Neural-Network (DNN) processing unit configured to:

execute, on image data, a part of a DNN algorithm by a first DNN;

generate a first result to be transmitted to an information processing device,

wherein the information processing device executes a remaining part of the DNN algorithm by a second DNN and generates second control information which indicates completion of the execution of the second DNN; and

receive the generated second control information from the information processing device.

11. The solid-state image capturing device according to claim 10 , wherein

the first result includes a feature map that is output from an intermediate layer of the DNN algorithm.

12. The solid-state image capturing device according to claim 10 , further comprising:

a storage that is configured to store therein at least the part of the DNN algorithm.

13. An information processing device, comprising:

a Deep-Neural-Network (DNN) processing unit configured to:

receive, from a solid-state image capturing device, a first result of execution of a part of a DNN algorithm by a first DNN on image data;

execute, on the first result, a remaining part of the DNN algorithm by a second DNN to generate a second result;

generate second control information which indicates completion of the execution of the second DNN; and

transmit the generated second control information to the solid-state image capturing device.

14. The information processing device according to claim 13 , further comprising:

a storage that is configured to store therein at least the remaining part of the DNN algorithm to be executed on the first result.

15. An image processing method, comprising:

in a solid-state image capturing device:

executing, on image data, a part of a Deep-Neural-Network (DNN) algorithm by a first DNN;

generating a first result to be transmitted to an information processing device,

wherein the information processing device executes a remaining part of the DNN algorithm by a second DNN and generates second control information which indicates completion of the execution of the second DNN; and

receiving the generated second control information from the information processing device.

16. An information processing method, comprising:

in an information processing device:

receiving, from a solid-state image capturing device, a first result of execution of a part of a Deep-Neural-Network (DNN) algorithm by a first DNN on image data;

executing, on the first result, a remaining part of the DNN algorithm by a second DNN to generate a second result;

generating second control information which indicates completion of the execution of the second DNN; and

transmitting the generated second control information to the solid-state image capturing device.

17. A non-transitory computer-readable medium having stored thereon, computer-executable instructions that when executed by a Deep-Neural-Network (DNN) processing unit, causes the DNN processing unit to execute operations, the operations comprising:

executing, on image data, a part of a DNN algorithm by a first DNN;

generating a first result to be transmitted to an information processing device,

wherein the information processing device executes a remaining part of the DNN algorithm by a second DNN and generates second control information which indicates completion of the execution of the second DNN; and

receiving the generated second control information from the information processing device.

18. A non-transitory computer-readable medium having stored thereon, computer-executable instructions that when executed by a Deep-Neural-Network (DNN) processing unit, causes the DNN processing unit to execute operations, the operations comprising:

receiving, from a solid-state image capturing device, a first result of execution of a part of a DNN algorithm by a first DNN on image data;

executing, on the first result, a remaining part of the DNN algorithm by a second DNN to generate a second result;

generating second control information which indicates completion of the execution of the second DNN; and

transmitting the generated second control information to the solid-state image capturing device.