Information processing apparatus, information processing method, and computer program

1. An information processing apparatus comprising:

plural information input units that input information including image information or sound information in an actual space;

an event detecting unit that generates event information including estimated position information and estimated identification information of users present in the actual space by analyzing the information inputted from the plural information input units, wherein:

the event detecting unit comprises a sound-event detecting unit and an image-event detecting unit;

the sound-event detecting unit generates identification information of users on the basis of comparison processing with characteristic information of user voices registered in advance; and

the image-event detecting unit generates user identification information on the basis of comparison processing with characteristic information of user faces registered in advance; and

an information-integration processing unit that:

sets probability distribution data of hypotheses concerning position and identification information of the users;

inputs the event information generated by the event detecting unit;

executes particle filtering processing to which plural particles set with plural targets corresponding to virtual users are applied;

executes update processing for the targets set for the respective particles taking into account elapsed time; and

executes generation of analysis information including user position information and user identification information of the users present in the actual space by updating and selecting the hypotheses on the basis of the event information.

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

the event detecting unit generates event information including estimated position information of users including a Gaussian distribution and user confidence factor information indicating probability values corresponding to the users, and

the information-integration processing unit holds particles set with plural targets having user position information including a Gaussian distribution corresponding to virtual users and user confidence factor information indicating probability values corresponding to the users, calculates event-target likelihoods that are indication values of similarities between the targets set for the respective particles and the event information, and executes particle setting processing with the target having a higher one of the event-target likelihoods preferentially set as an event occurrence source hypothesis target.

3. An information processing apparatus according to claim 2 , wherein the information-integration processing unit calculates sum data of the event-target likelihoods and particle weights set for the respective particles as target weights and executes particle setting processing with a target having a large target weight preferentially set as an event occurrence source hypothesis target.

4. An information processing apparatus according to claim 2 , wherein the information-integration processing unit calculates likelihoods between event occurrence source hypothesis targets set for the respective particles and the event information inputted from the event detecting unit and sets values corresponding to levels of the likelihoods for the respective particles as particle weights.

5. An information processing apparatus according to claim 4 , wherein the information-integration processing unit executes re-sampling processing for preferentially reselecting the particle having a larger one of the particle weights and performs update processing for the particle.

6. An information processing apparatus according to claim 2 , wherein the information-integration processing unit performs, concerning event occurrence source hypothesis targets set for the respective particles, update processing to which the event information generated by the event detecting unit is applied.

7. An information processing apparatus according to claim 4 , wherein the information-integration processing unit generates target information in which an integrated sum of target data and the particle weights set for the respective particles is set as user position information and user identification information corresponding to the respective targets.

8. An information processing apparatus according to claim 2 , wherein the information-integration processing unit generates signal information as a probability value of an event occurrence source according to numbers of event occurrence source hypothesis targets set for the respective particles.

9. An information processing apparatus according to claim 2 , wherein the information-integration processing unit generates a provisional target having user position information and user identification information including uniform data and performs, when a likelihood between the provisional target and the event information generated by the event detecting unit is a value larger than an event-target likelihood corresponding to an existing target, processing for adding the provisional target to the respective particles anew.

10. An information processing apparatus according to claim 4 , wherein the information-integration processing unit executes, when a peak value of Gaussian distribution data as user position information included in an integrated sum of target data and the particle weights set for the respective particles is smaller than a threshold set in advance, processing for deleting the target.

11. An information processing method for executing information analysis processing in an information processing apparatus, the information processing method comprising:

an information inputting step in which plural information input units input information including image information or sound information in an actual space;

an event detecting step in which an event detecting unit generates event information including estimated position information and estimated identification information of users present in the actual space by analyzing the information inputted in the information inputting step, wherein:

the event detecting unit comprises a sound-event detecting unit and an image-event detecting unit;

the sound-event detecting unit generates identification information of users on the basis of comparison processing with characteristic information of user voices registered in advance; and

the image-event detecting unit generates user identification information on the basis of comparison processing with characteristic information of user faces registered in advance; and

an information-integration processing step in which an information-integration processing unit:

sets probability distribution data of hypotheses concerning position and identification information of the users;

inputs the event information generated by the event detecting unit;

executes particle filtering processing to which plural particles set with plural targets corresponding to virtual users are applied;

executes update processing for the targets set for the respective particles taking into account elapsed time; and

executes generation of analysis information including user position information and user identification information of the users present in the actual space by updating and selecting the hypotheses on the basis of the event information.

12. An information processing method according to claim 11 , wherein

the event detecting step is a step of generating event information including estimated position information of users including a Gaussian distribution and user confidence factor information indicating probability values corresponding to the users,

the information-integration processing unit holds particles set with plural targets having user position information including a Gaussian distribution corresponding to virtual users and user confidence factor information indicating probability values corresponding to the users, and

the information-integration processing step is a step of calculating event-target likelihoods that are indication values of similarities between the targets set for the respective particles and the event information and executing particle setting processing with the target having a higher one of the event-target likelihoods preferentially set as an event occurrence source hypothesis target.

13. An information processing method according to claim 12 , wherein the information-integration processing step is a step of calculating sum data of the event-target likelihoods and particle weights set for the respective particles as target weights and executing particle setting processing with a target having a large target weight preferentially set as an event occurrence source hypothesis target.

14. An information processing method according to claim 12 , wherein the information-integration processing step is a step of calculating likelihoods between event occurrence source hypothesis targets set for the respective particles and the event information inputted from the event detecting unit and setting values corresponding to levels of the likelihoods for the respective particles as particle weights.

15. An information processing method according to claim 14 , wherein the information-integration processing step is a step of executing re-sampling processing for preferentially reselecting the particle having a larger one of the particle weights and performing update processing for the particle.

16. An information processing method according to claim 12 , wherein the information-integration processing step is a step of performing, concerning event occurrence source hypothesis targets set for the respective particles, update processing to which the event information generated by the event detecting unit is applied.

17. An information processing method according to claim 14 , wherein the information-integration processing step is a step of generating target information in which an integrated sum of target data and the particle weights set for the respective particles is set as user position information and user identification information corresponding to the respective targets.

18. An information processing method according to claim 12 , wherein the information-integration processing step is a step of generating signal information as a probability value of an event occurrence source according to numbers of event occurrence source hypothesis targets set for the respective particles.

19. An information processing method according to claim 12 , wherein the information-integration processing step includes a step of generating a provisional target having user position information and user identification information including uniform data and performing, when a likelihood between the provisional target and the event information generated by the event detecting unit is a value larger than an event-target likelihood corresponding to an existing target, processing for adding the provisional target to the respective particles anew.

20. An information processing method according to claim 14 , wherein the information-integration processing step includes a step of executing, when a peak value of Gaussian distribution data as user position information included in an integrated sum of target data and the particle weights set for the respective particles is smaller than a threshold set in advance, processing for deleting the target.

21. A non-transitory computer-readable medium for causing an information processing apparatus to execute information analysis processing, the non-transitory computer-readable medium causing the information processing apparatus to execute:

an information inputting step of causing plural information input units to input information including image information or sound information in an actual space;

an event detecting step of causing an event detecting unit to generate event information including estimated position information and estimated identification information of users present in the actual space by analyzing the information inputted in the information inputting step, wherein:

the event detecting unit comprises a sound-event detecting unit and an image-event detecting unit;

the sound-event detecting unit generates identification information of users on the basis of comparison processing with characteristic information of user voices registered in advance; and

the image-event detecting unit generates user identification information on the basis of comparison processing with characteristic information of user faces registered in advance; and

an information-integration processing step of causing an information-integration processing unit to:

set probability distribution data of hypotheses concerning position and identification information of the users;

input the event information generated by the event detecting unit;

execute particle filtering processing to which plural particles set with plural targets corresponding to virtual users are applied;

execute update processing for the targets set for the respective particles taking into account elapsed time; and

execute generation of analysis information including user position information and user identification information of the users present in the actual space by updating and selecting the hypotheses on the basis of the event information.