PRIVACY-PRESERVING BEHAVIOR DETECTION

• US 20190096209A1
• Filed: 09/22/2017
• Published: 03/28/2019
• Est. Priority Date: 09/22/2017
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Chen, Ke-Yu, Yang, Shao-Wen, Wan, Chieh-Yih, Shah, Rahul C., Sanjay, Addicam V., Thomas, Scott A.
• Granted Patent: US 10,467,873 B2
• Time in Patent Office: Days
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G08B 13/19604 involving reference image or background adaptation with time to compensate for changing conditions, e.g. reference image update on detection of light level change
• CPC Class Codes: G08B 13/2454 Checking of authorisation of a person accessing tagged items in an EAS system
• CPC Class Codes: H04W 4/023 using mutual or relative location information between multiple location based services [LBS] targets or of distance thresholds
• CPC Class Codes: H04W 4/80 Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication

Abstract

In one embodiment, an apparatus may comprise a sensor to detect a plurality of radio signals from one or more transmitters. The apparatus may further comprise a processor to: identify the plurality of radio signals detected by the sensor; detect a proximity of one or more assets based on the plurality of radio signals, wherein the one or more assets are associated with the one or more transmitters; identify the one or more assets based on an identity of the one or more transmitters, wherein each transmitter is associated with a particular asset; identify a plurality of signal characteristics associated with the plurality of radio signals; detect a proximity of a human based on the plurality of signal characteristics; and detect one or more human-asset interactions based on the plurality of signal characteristics.

Claims

1. An apparatus, comprising:

a sensor to detect a plurality of radio signals from one or more transmitters; and

a processor to;

identify the plurality of radio signals detected by the sensor;

detect a proximity of one or more assets based on the plurality of radio signals, wherein the one or more assets are associated with the one or more transmitters;

identify the one or more assets based on an identity of the one or more transmitters, wherein each transmitter is associated with a particular asset;

identify a plurality of signal characteristics associated with the plurality of radio signals;

detect a proximity of a human based on the plurality of signal characteristics; and

detect one or more human-asset interactions based on the plurality of signal characteristics, wherein the one or more human-asset interactions are detected based on a machine learning model, wherein the machine learning model is trained to recognize different human-asset interactions based on the plurality of signal characteristics.

2. The apparatus of claim 1, wherein the sensor comprises an RFID reader, and wherein the one or more transmitters comprise one or more RFID tags.

3. The apparatus of claim 1, wherein the plurality of signal characteristics comprises:

signal phase information;

signal strength information;

or

signal frequency information.

4. The apparatus of claim 1, wherein the processor to detect the proximity of the human based on the plurality of signal characteristics is further to:

identify a fluctuation in the plurality of signal characteristics; and

determine that the fluctuation is indicative of signal interference associated with the proximity of the human.

5. The apparatus of claim 1, wherein the processor to detect the one or more human-asset interactions based on the plurality of signal characteristics is further to:

identify a fluctuation in the plurality of signal characteristics;

determine that the fluctuation is indicative of movement from the plurality of assets; and

identify the one or more human-asset interactions based on the movement from the plurality of assets.

7. The apparatus of claim 1, wherein the processor is further to:

detect a proximity of a mobile device associated with the human; and

determine an identity of the mobile device.

8. The apparatus of claim 7, wherein the processor is further to identify the human based on the identity of the mobile device.

9. The apparatus of claim 7, wherein the processor is further to obtain information associated with the human based on the identity of the mobile device.

10. The apparatus of claim 9, wherein the information associated with the human comprises customer sales information.

26. The apparatus of claim 1, wherein:

the one or more human-asset interactions comprise a plurality of interactions between a retail customer and a plurality of clothing items; and

the processor to detect the one or more human-asset interactions based on the plurality of signal characteristics is further to;

determine which of the plurality of clothing items are tried on by the retail customer;

determine an order in which the plurality of clothing items are tried on by the retail customer;

determine one or more groupings of the plurality of clothing items that are tried on together by the retail customer;

determine an amount of time in which each of the plurality of clothing items is tried on by the retail customer;

or

determine a number of times in which each of the plurality of clothing items is tried on by the retail customer.

27. The apparatus of claim 7, wherein:

the one or more human-asset interactions comprise a plurality of customer-product interactions between a retail customer and a plurality of retail products; and

the processor is further to;

identify one or more actual sales associated with the retail customer based on the identity of the mobile device; and

train a machine learning model to identify one or more correlations between the one or more actual sales and the plurality of customer-product interactions.

6. (canceled)

11. At least one non-transitory machine accessible storage medium having instructions stored thereon, wherein the instructions, when executed on a machine, cause the machine to:

identify a plurality of radio signals received by a sensor from one or more transmitters;

detect a proximity of one or more assets based on the plurality of radio signals, wherein the one or more assets are associated with the one or more transmitters;

identify the one or more assets based on an identity of the one or more transmitters, wherein each transmitter is associated with a particular asset;

identify a plurality of signal characteristics associated with the plurality of radio signals;

detect a proximity of a human based on the plurality of signal characteristics; and

detect one or more human-asset interactions based on the plurality of signal characteristics, wherein the one or more human-asset interactions are detected based on a machine learning model, wherein the machine learning model is trained to recognize different human-asset interactions based on the plurality of signal characteristics.

12. The storage medium of claim 11, wherein the sensor comprises an RFID reader, and wherein the one or more transmitters comprise one or more RFID tags.

13. The storage medium of claim 11, wherein the plurality of signal characteristics comprises:

signal phase information;

signal strength information;

or

signal frequency information.

14. The storage medium of claim 11, wherein the instructions that cause the machine to detect the proximity of the human based on the plurality of signal characteristics further cause the machine to:

identify a fluctuation in the plurality of signal characteristics; and

determine that the fluctuation is indicative of signal interference associated with the proximity of the human.

15. The storage medium of claim 11, wherein the instructions that cause the machine to detect the one or more human-asset interactions based on the plurality of signal characteristics further cause the machine to:

identify a fluctuation in the plurality of signal characteristics;

determine that the fluctuation is indicative of movement from the plurality of assets; and

identify the one or more human-asset interactions based on the movement from the plurality of assets.

17. The storage medium of claim 11, wherein the instructions further cause the machine to:

detect a proximity of a mobile device associated with the human; and

determine an identity of the mobile device.

18. The storage medium of claim 17, wherein the instructions further cause the machine to:

identify the human based on the identity of the mobile device; and

obtain information associated with the human based on the identity of the mobile device.

16. (canceled)

19. A system, comprising:

one or more transmitters to transmit a plurality of radio signals, wherein the one or more transmitters are associated with one or more assets;

one or more sensors to detect the plurality of radio signals from the one or more transmitters; and

a processor to;

identify the plurality of radio signals detected by the one or more sensors;

detect a proximity of the one or more assets based on the plurality of radio signals;

identify the one or more assets based on an identity of the one or more transmitters, wherein each transmitter is associated with a particular asset;

identify a plurality of signal characteristics associated with the plurality of radio signals;

detect a proximity of a human based on the plurality of signal characteristics; and

detect one or more human-asset interactions based on the plurality of signal characteristics, wherein the one or more human-asset interactions are detected based on a machine learning model, wherein the machine learning model is trained to recognize different human-asset interactions based on the plurality of signal characteristics.

20. The system of claim 19, wherein the one or more transmitters comprise one or more RFID tags, and wherein the one or more sensors comprise one or more RFID readers.

21. The system of claim 19, further comprising:

a database to store a representation of the one or more human-asset interactions; and

a gateway in communication with the one or more sensors, wherein the gateway is to transmit the representation of the one or more human-asset interactions to a data processing service.

22. The system of claim 19, wherein:

the one or more assets comprise one or more retail products; and

the one or more human-asset interactions comprise one or more customer-product interactions.

23. A method, comprising:

identifying a plurality of radio signals received by a sensor from one or more transmitters;

detecting a proximity of one or more assets based on the plurality of radio signals, wherein the one or more assets are associated with the one or more transmitters;

identifying the one or more assets based on an identity of the one or more transmitters, wherein each transmitter is associated with a particular asset;

identifying a plurality of signal characteristics associated with the plurality of radio signals;

detecting a proximity of a human based on the plurality of signal characteristics; and

detecting one or more human-asset interactions based on the plurality of signal characteristics, wherein the one or more human-asset interactions are detected based on a machine learning model, wherein the machine learning model is trained to recognize different human-asset interactions based on the plurality of signal characteristics.

24. The method of claim 23, wherein detecting the one or more human-asset interactions based on the plurality of signal characteristics comprises:

identifying a fluctuation in the plurality of signal characteristics;

determining that the fluctuation is indicative of movement from the plurality of assets; and

identifying the one or more human-asset interactions based on the movement from the plurality of assets.

25. (canceled)