Information processing apparatus, information processing method, and program

1. An information processing apparatus comprising:

at least one memory configured to store instructions; and

at least one processor configured to execute the instructions to perform operations, the operations comprising:

detecting a specific event from video data;

receiving, from a user, input for specifying processing to be executed; and

executing first processing specified by the input, and executing second processing of generating learning data used for machine learning and storing the generated learning data in a storage apparatus, wherein

the second processing includes

determining, based on a classification of the first processing specified by the input, whether a detection result of the specific event is correct, and

generating the learning data including at least a part of the video data, category information indicating a category of the detected specific event, and correct/incorrect information indicating whether the detection result of the specific event is correct or incorrect, wherein

the first processing is classified into either positive processing executed by assuming that the detection result of the specific event is correct or negative processing executed by assuming the detection result of the specific event is incorrect, and

the operations comprise, in the second processing:

determining, when the first processing specified by the input is classified into the positive processing, that the detection result of the specific event is correct, and

determining, when the first processing specified by the input is classified into the negative processing, that the detection result of the specific event is incorrect.

2. The information processing apparatus according to claim 1 , wherein

the positive processing includes at least any one of report processing to a predetermined address, processing of starting an alarm apparatus, processing of modifying a display area of the video data, processing of selecting a partial area of the video data, processing of starting recording the video data, and processing of stopping recording the video data and storing recorded data, and

the negative processing includes at least any one of processing of returning to a mode for detecting the specific event, and processing of stopping recording the video data and discarding recorded data.

3. The information processing apparatus according to claim 1 , wherein the operations further comprise:

detecting, as the specific event, at least any one of a person, an object, and a behavior of a person or an object, matched with a predetermined condition.

4. The information processing apparatus according to claim 1 , wherein

the second processing includes:

acquiring area information indicating an area where the specific event is detected in the video data, and

further including the area information in the learning data.

5. The information processing apparatus according to claim 4 , wherein

the operations further comprise:

executing processing of correcting an area where the specific event is detected, and

acquiring the area information with respect to the corrected area.

6. An information processing apparatus comprising:

at least one memory configured to store instructions; and

at least one processor configured to execute the instructions to perform operations, the operations comprising:

detecting one or more specific events from video data;

receiving, from a user, input for specifying processing to be executed; and

executing first processing specified by the input, and executing second processing of generating learning data used for machine learning and storing the generated learning data in a storage apparatus, wherein

the second processing includes:

determining, based on a classification of the first processing specified by the input, whether a detection result of the specific event is correct, and

generating the learning data including at least a part of the video data, category information indicating a category of each detected specific event, and correct/incorrect information indicating whether the detection result of each specific event is correct or incorrect, wherein the operations further comprise:

when two or more specific events different in category from each other are detected,

further receiving in addition to the input for specifying the processing to be executed, second input for specifying the category of each specific event, and

determining, in the second processing, based on the category specified by the second input, the category information to be included in the learning data.

7. An information processing method executed by a computer, comprising:

detecting a specific event from video data;

receiving, from a user, input for specifying processing to be executed; and

executing first processing specified by the input, and executing second processing of generating learning data used for machine learning and storing the generated learning data in a storage apparatus, wherein

the second processing includes:

determining, based on a classification of the first processing specified by the input, whether a detection result of the specific event is correct, and

generating the learning data including at least a part of the video data, category information indicating a category of the detected specific event, and correct/incorrect information indicating whether the detection result of the specific event is correct or incorrect, wherein

the first processing is classified into either positive processing executed by assuming that the detection result of the specific event is correct or negative processing executed by assuming the detection result of the specific event is incorrect, and

the computer, in the second processing,

determines, when the first processing specified by the input is classified into the positive processing, that the detection result of the specific event is correct, and

determines, when the first processing specified by the input is classified into the negative processing, that the detection result of the specific event is incorrect.

8. The information processing method according to claim 7 , wherein

the positive processing includes at least any one of report processing to a predetermined address, processing of starting an alarm apparatus, processing of modifying a display area of the video data, processing of selecting a partial area of the video data, processing of starting recording the video data, and processing of stopping recording the video data and storing recorded data, and

the negative processing includes at least any one of processing of returning to a mode for detecting the specific event, and processing of stopping recording the video data and discarding recorded data.

9. The information processing method according to claim 7 , wherein

the computer detects, as the specific event, at least any one of a person, an object, and a behavior of a person or an object, matched with a predetermined condition.

10. The information processing method according to claim 7 , wherein

the computer, in the second processing,

acquires area information indicating an area where the specific event is detected in the video data, and

further includes the area information in the learning data.

11. The information processing method according to claim 10 , wherein

the computer

executes processing of correcting an area where the specific event is detected, and

acquires the area information with respect to the corrected area.

12. An information processing method executed by a computer, comprising:

detecting one or more specific events from video data;

receiving, from a user, input for specifying processing to be executed; and

executing first processing specified by the input, and executing second processing of generating learning data used for machine learning and storing the generated learning data in a storage apparatus, wherein

the second processing includes:

determining, based on a classification of the first processing specified by the input, whether a detection result of each specific event is correct, and

generating the learning data including at least a part of the video data, category information indicating a category of each detected specific event, and correct/incorrect information indicating whether the detection result of each specific event is correct or incorrect, wherein

the computer,

when two or more specific events different in category from each other are detected,

further receives, in addition to the input for specifying the processing to be executed, second input for specifying the category of each specific event, and

determines, in the second processing, based on the category specified by the second input, the category information to be included in the learning data.

13. A non-transitory computer readable medium storing a program causing a computer to execute an information processing method, the method comprising:

detecting a specific event from video data;

receiving, from a user, input for specifying processing to be executed; and

executing first processing specified by the input, and executing second processing of generating learning data used for machine learning and storing the generated learning data in a storage apparatus, wherein

the second processing includes:

determining, based on a classification of the first processing specified by the input, whether a detection result of the specific event is correct, and

generating the learning data including at least a part of the video data, category information indicating a category of the detected specific event, and correct/incorrect information indicating whether the detection result of the specific event is correct or incorrect wherein

the first processing is classified into either positive processing executed by assuming that the detection result of the specific event is correct or negative processing executed by assuming the detection result of the specific event is incorrect, and

the computer, in the second processing,

determines, when the first processing specified by the input is classified into the positive processing, that the detection result of the specific event is correct, and

determines, when the first processing specified by the input is classified into the negative processing, that the detection result of the specific event is incorrect.

14. A non-transitory computer readable medium storing a program causing a computer to execute an information processing method, the method comprising:

detecting one or more specific events from video data;

receiving, from a user, input for specifying processing to be executed; and

executing first processing specified by the input, and executing second processing of generating learning data used for machine learning and storing the generated learning data in a storage apparatus, wherein

the second processing includes:

determining, based on a classification of the first processing specified by the input, whether a detection result of each specific event is correct, and

generating the learning data including at least a part of the video data, category information indicating a category of each detected specific event, and correct/incorrect information indicating whether the detection result of each specific event is correct or incorrect, wherein

the computer,

when two or more specific events different in category from each other are detected,

further receives, in addition to the input for specifying the processing to be executed, second input for specifying the category of each specific event, and

determines, in the second processing, based on the category specified by the second input, the category information to be included in the learning data.