Creating a training tool

• US 8,612,286 B2
• Filed: 10/31/2008
• Issued: 12/17/2013
• Est. Priority Date: 10/31/2008
• Status: Expired due to Fees

• Current Assignee: Kyndryl Incorporated (Kyndryl Holdings, Inc.)
• Original Assignee: International Business Machines Corporation
• Inventors: Bobbitt, Russell Patrick, Fan, Quanfu, Hampapur, Arun, Kjeldsen, Frederik, Pankanti, Sharathchandra Umapathirao, Yanagawa, Akira, Zhai, Yun
• Primary Examiner(s): Sterrett, Jonathan G
• Application Number:
  US12/262,467
• Publication Number:
  US 20100114671A1
• Time in Patent Office: 1,873 Days
• Field of Search: 705/7.42, 348/150, 382/155
• US Class Current: 705/7.42
• CPC Class Codes: G06Q 10/063114 Status monitoring or status determination for a person or group
• CPC Class Codes: G07G 3/00 Alarm indicators, e.g. bells
• CPC Class Codes: G07G 3/006 False operation
• CPC Class Codes: G09B 19/18 Book-keeping or economics
• CPC Class Codes: H04N 7/18 Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast

Abstract

Techniques for creating a training technique for an individual are provided. The techniques include obtaining video of one or more events and information from a transaction log that corresponds to the one or more events, wherein the one or more events relate to one or more actions of an individual, classifying the one or more events into one or more event categories, comparing the one or more classified events with an enterprise best practices model to determine a degree of compliance, examining the one or more classified events to correct one or more misclassifications, if any, and revise the one or more event categories with the one or more corrected misclassifications, if any, and using the degree of compliance to create a training technique for the individual.

Claims

1. A method for creating a training technique for an individual at a retail checkout environment, comprising the steps of:

building an event model for a retail checkout environment, wherein said building an event model comprises;

iteratively capturing a video frame via a camera associated with the retail checkout environment and importing a concurrent record of a transaction event occurring at the retail checkout environment; and

corresponding the record of the transaction event with one or more features of the captured video frame to identify the transaction event within the event model;

obtaining video of one or more events at a retail checkout environment and information from a transaction log that corresponds to the one or more events, wherein one or more events relate to one or more actions of an individual, and wherein obtaining video of one or more events and information from a transaction log that corresponds to the one or more events is carried out by a video analytics engine executing on a hardware processor;

automatically classifying each of the one or more events into an event category by comparing the obtained video and information from the transaction log to the event model using a classification technique, wherein event categories include a person present in cashier area, a barcode scanned, multiple people present in customer area, transaction voided, a person present in customer area, multiple people present in cashier area, a keyboard interaction, one or more items bagged, a pick-up motion, a scan motion and a drop motion, and wherein automatically classifying the one or more events is carried out by an event classifier executing on a hardware processor;

examining the one or more classified events to correct one or more misclassifications, if any, and revise the one or more event categories with the one or more corrected misclassifications, if any; and

automatically generating a report that identifies each classified event and one or more enterprise rules corresponding to the identified classified event wherein generating a report is carried out by a compliance engine executing on a hardware processor.

2. The method of claim 1, wherein obtaining video of one or more events and information from a transaction log that corresponds to the one or more events comprises inputting the video and transaction log information into a video analytics engine, wherein the video analytics engine outputs a set of one or more unclassified events.

3. The method of claim 1, wherein each of the one or more events comprises a collection of one or more features.

4. The method of claim 3, wherein the one or more features comprise at least one of shape, color, texture, location, orientation, area, one or more motion characteristics, one or more edges, optical flow, one or more color statistics, spatial gradient, temporal gradient, temporal texture, one or more object locations and one or more object trajectories.

5. The method of claim 1, further comprising generating a compliance report that indicates the degree of compliance for each individual.

6. The method of claim 1, wherein using the degree of compliance to create a training technique for the individual comprises a human supervisor examining the degree of compliance to create a training technique for the individual.

7. The method of claim 1, further comprising rating the classification of the one or more events based on a metric.

8. The method of claim 1, further comprising automatically learning one or more statistical models of one or more model events.

9. The method of claim 1, further comprising generating a revised transaction log.

10. A computer program product comprising a tangible computer readable recordable storage device having computer readable program code for creating a training technique for an individual at a retail checkout environment, said computer program product including:

computer readable program code for building an event model for a retail checkout environment, wherein said building an event model comprises;

iteratively capturing a video frame via a camera associated with the retail checkout environment and importing a concurrent record of a transaction event occurring at the retail checkout environment; and

corresponding the record of the transaction event with one or more features of the captured video frame to identify the transaction event within the event model;

computer readable program code for obtaining video of one or more events at a retail checkout environment and information from a transaction log that corresponds to the one or more events, wherein one or more events relate to one or more actions of an individual;

computer readable program code for automatically classifying each of the one or more events into an event category by comparing the obtained video and information from the transaction log to the event model using a classification technique, wherein event categories include a person present in cashier area, a barcode scanned, multiple people present in customer area, transaction voided, a person present in customer area, multiple people present in cashier area, a keyboard interaction, one or more items bagged, a pick-up motion, a scan motion and a drop motion;

computer readable program for examining the one or more classified events to correct one or more misclassifications, if any, and revise the one or more event categories with the one or more corrected misclassifications, if any; and

computer readable program automatically generating a report that identifies each classified event and one or more enterprise rules corresponding to the identified classified event.

11. The computer program product of claim 10, wherein the computer readable program code for obtaining video of one or more events and information from a transaction log that corresponds to the one or more events comprises computer readable program code for inputting the video and transaction log information into a video analytics engine, wherein the video analytics engine outputs a set of one or more unclassified events.

12. The computer program product of claim 10, further comprising computer readable program code for generating a compliance report that indicates the degree of compliance for each individual.

13. The computer program product of claim 10, wherein the computer readable program code for using the degree of compliance to create a training technique for the individual comprises computer readable program code for a human supervisor examining the degree of compliance to create a training technique for the individual.

14. The computer program product of claim 10, further comprising computer readable program code for rating the classification of the one or more events based on a metric.

15. The computer program product of claim 10, further comprising computer readable program code for generating a revised transaction log.

16. The computer program product of claim 10, further comprising computer readable program code for automatically learning one or more statistical models of one or more model events.

17. A system for creating a training technique for an individual at a retail checkout environment, comprising:

a memory; and

at least one processor coupled to said memory and operative to;

build an event model for a retail checkout environment, wherein said building an event model comprises;

iteratively capturing a video frame via a camera associated with the retail checkout environment and importing a concurrent record of a transaction event occurring at the retail checkout environment; and

correspond the record of the transaction event with one or more features of the captured video frame to identify the transaction event within the event model;

obtain video of one or more events at a retail checkout environment and information from a transaction log that corresponds to the one or more events, wherein one or more events relate to one or more actions of an individual;

automatically classify each of the one or more events into an event category by comparing the obtained video and information from the transaction log to the event model using a classification technique, wherein event categories include a person present in cashier area, a barcode scanned, multiple people present in customer area, transaction voided, a person present in customer area, multiple people present in cashier area, a keyboard interaction, one or more items bagged, a pick-up motion, a scan motion and a drop motion;

examine the one or more classified events to correct one or more misclassifications, if any, and revise the one or more event categories with the one or more corrected misclassifications, if any; and

automatically generate a report that identifies each classified event and one or more enterprise rules corresponding to the identified classified event.

18. The system of claim 17, wherein in obtaining video of one or more events and information from a transaction log that corresponds to the one or more events, the at least one processor coupled to said memory is further operative to input the video and transaction log information into a video analytics engine, wherein the video analytics engine outputs a set of one or more unclassified events.

19. The system of claim 17, wherein the at least one processor coupled to said memory is further operative to generate a compliance report that indicates the degree of compliance for each individual.

20. The system of claim 17, wherein in using the degree of compliance to create a training technique for the individual, the at least one processor coupled to said memory is further operative to enable a human supervisor examining the degree of compliance to create a training technique for the individual.

21. The system of claim 17, wherein the at least one processor coupled to said memory is further operative to rate the classification of the one or more events based on a metric.

22. The system of claim 17, wherein the at least one processor coupled to said memory is further operative to generate a revised transaction log.

23. The system of claim 17, wherein the at least one processor coupled to said memory is further operative to automatically learn one or more statistical models of one or more model events.

24. An apparatus for creating a training technique for an individual at a retail checkout environment, said apparatus comprising:

means for building an event model for a retail checkout environment, said means comprising a module executing on a hardware processor, and wherein said building an event model comprises;

iteratively capturing a video frame via a camera associated with the retail checkout environment and importing a concurrent record of a transaction event occurring at the retail checkout environment; and

corresponding the record of the transaction event with one or more features of the captured video frame to identify the transaction event within the event model;

means for obtaining video of one or more events at a retail checkout environment and information from a transaction log that corresponds to the one or more events, said means comprising a module executing on a hardware processor, and wherein one or more events relate to one or more actions of an individual;

means for automatically classifying each of the one or more events into an event category by comparing the obtained video and information from the transaction log to the event model using a classification technique, said means comprising a module executing on a hardware processor, and wherein event categories include a person present in cashier area, a barcode scanned, multiple people present in customer area, transaction voided, a person present in customer area, multiple people present in cashier area, a keyboard interaction, one or more items bagged, a pick-up motion, a scan motion and a drop motion;

means for examining the one or more classified events to correct one or more misclassifications, if any, and revise the one or more event categories with the one or more corrected misclassifications, if any, said means comprising a module executing on a hardware processor; and

means for automatically generating a report that identifies each classified event and one or more enterprise rules corresponding to the identified classified event, said means comprising a module executing on a hardware processor.