Directional impression analysis using deep learning
Systems and techniques are provided for detecting gaze direction of subjects in an area of real space. The system receives a plurality of sequences of frames of corresponding fields of view in the real space. The system uses sequences of frames in a plurality of sequences of frames to identify locations of an identified subject and gaze directions of the subject in the area of real space over time. The system includes logic having access to a database identifying locations of items in the area of real space. The system identifies items in the area of real space matching the identified gaze directions of the identified subject.
1. A method including:
using sequences of frames produced by sensors to identify (i) locations of identified subjects in an area of real space and (ii) interest, by the identified subjects, in identified items positioned at locations in the area of real space over time;
storing the identified items for which interest was identified as entries in a database, the entries including an item identifier and a location of the items in the area of real space; and
generating a heat map of at least a portion of the area of real space, wherein the heat map identifies locations, in the area of real space, of identified items for which the interest by the identified subjects exceeds a threshold.
2. The method of claim 1 , further including calculating a unique visitor directional impression from the entries in the database, wherein the unique visitor directional impression identifies a number of unique or distinct subjects who have an interest in identified items.
3. The method of claim 2 , further including calculating unique visitor directional impressions over a period of time.
4. The method of claim 2 , further including a determining length of time for which an identified subject has interest in a particular identified item and storing the determined length of time in an entry in the entries in the database.
5. The method of claim 4 , further including calculating a length of time for which a particular identified item is of interest to a particular identified subject during the particular identified subject's presence in the area of real space.
6. The method of claim 5 , further including determining a particular location in the area of real space at which a particular item is of interest to the particular identified subject and displaying a view of the particular location.
7. The method of claim 4 , further including calculating a distance of an identified subject from an item positioned at a location in the area of real space, the distance being used to identify an interest of the identified subject in the item.
8. The method of claim 7 , wherein the calculating of the distance of the identified subject from the item uses a location of a neck joint of the identified subject and a location of an inventory display structure in which the item is positioned.
9. The method of claim 1 , further including updating a store planogram using the heat map of the area of real space and using the updated store planogram to facilitate stocking of items in the area of real space.
10. A system including one or more processors and a memory, the memory comprising instructions that, when executed, cause the one or more processors to perform operations comprising:
using sequences of frames produced by sensors to identify (i) locations of identified subjects in an area of real space and (ii) interest, by the identified subjects, in identified items positioned at locations in the area of real space;
storing the identified items for which interest was identified as entries in a database, the entries including an item identifier and a location of the items in the area of real space; and
generating a heat map of at least a portion of the area of real space, wherein the heat map identifies locations, in the area of real space, of identified items for which the interest by the identified subjects exceeds a threshold.
11. The system of claim 10 , wherein the operations further include calculating a unique visitor directional impression from the entries in the database, wherein the unique visitor directional impression identifies a number of unique or distinct subjects who have in interest in identified items.
12. The system of claim 11 , wherein the operations further include calculating unique visitor directional impressions over a period of time.
13. The system of claim 11 , wherein the operations further include determining a length of time for which an identified subject has interest in a particular identified item and storing the determined length of time in an entry in the entries.
14. The system of claim 13 , wherein the operations further include calculating a length of time for which a particular identified item is of interest to a particular identified subject during the particular identified subject's presence in the area of real space.
15. The system of claim 14 , wherein the operations further includes determining a particular location in the area of real space at which a particular item is of interest to the particular identified subject and displays a view of the particular location.
16. The system of claim 13 , wherein the operations further includes calculating a distance of an identified subject from an item positioned at a location in the area of real space, the distance being used to identify an interest of the identified subject in the item.
17. The system of claim 16 , wherein the calculating of the distance uses a location of a neck joint of the identified subject and a location of an inventory display structure in which the item is positioned.
18. The system of claim 10 , wherein the operations further includes updating a store planogram using the heat map of the area of real space and use the updated store planogram to facilitate stocking of items in the area of real space.
19. A non-transitory computer readable storage medium impressed with computer program instructions to detect gaze direction of subjects in an area of real space, the instructions, when executed on a processor, implement a method comprising:
using sequences of frames produced by sensors to identify (i) locations of identified subjects in an area of real space and (ii) interest, by the identified subjects, in identified items positioned at locations in the area of real space over time;
storing the identified items for which interest was identified as entries in a database, the entries including an item identifier and a location of the items in the area of real space; and
generating a heat map of at least a portion of the area of real space, wherein the heat map identifies locations, in the area of real space, of identified items for which the interest by the identified subjects exceeds a threshold.
20. The non-transitory computer readable storage medium of claim 19 , wherein the method further includes calculating a unique visitor directional impression from the entries in the database, wherein the unique visitor directional impression identifies a number of unique or distinct subjects who have an interest in identified items.