Picture ranking method, and terminal

Patent metadata

Field	Value
Publication number	US-10755086-B2
Application number	US-201916276206-A
Country	US
Kind code	B2
Filing date	Feb 14, 2019
Priority date	Mar 12, 2014
Publication date	Aug 25, 2020
Grant date	Aug 25, 2020

Abstract

A picture ranking method and a terminal comprises acquiring pictures stored in a terminal, detecting whether the pictures are first-type pictures, where a first-type picture refers to a picture including a human face, and when the pictures are first-type pictures, ranking the pictures according to a social relation model, or when the pictures are not first-type pictures, ranking the pictures according to a preset rule.

First claim

What is claimed is:

1. A picture ranking method, comprising:
acquiring a user's pictures;
detecting first-type pictures from the user's pictures, wherein the first-type pictures comprise one or more human faces;
determining contacts associated with each of the first-type pictures according to the one or more human faces in the first-type pictures;
determining social relations between the contacts associated with each of the first-type pictures and the user based on an identity feature of each of the contacts, wherein the identity feature of a contact describes a type of relationship between the user and each of the contacts;
ranking the first-type pictures according to the social relations; and
presenting the first-type pictures after ranking the first-type pictures.

2. The picture ranking method of claim 1 , wherein determining the social relations between the contacts associated with each of the first-type pictures and the user comprises determine the social relations between the contacts associated with each of the first-type pictures and the user according to a social relation model.

3. The picture ranking method of claim 2 , wherein determining the social relations according to the social relation model comprises
acquiring information about the contacts associated with each of the first-type pictures using the social relation model, wherein the information about the contacts associated with each of the first-type pictures comprises the identity feature of each of the contacts.

4. The picture ranking method of claim 2 , wherein before determining the social relations according to the social relation model, the method further comprises training the social relation model using a stored profile picture of the contacts and information about the contacts.

5. The picture ranking method of claim 4 , wherein training the social relation model using the stored profile picture of the contacts and the information about the contacts comprises:
extracting profile picture data of the contacts; and
annotating the profile picture data using the information about the contacts to obtain the social relation model, wherein the information about the contacts comprises the identity feature of each of the contacts.

6. The picture ranking method of claim 1 , wherein detecting the first-type pictures from the user's pictures comprises detecting the first-type pictures according to a facial recognition model.

7. The picture ranking method of claim 6 , wherein detecting the first-type pictures comprises according to the facial recognition model:
extracting feature values of the user's pictures;
performing a matching between the feature values of the user's pictures and the facial recognition model; and
detecting the first-type pictures based on the first-type pictures that are higher than a first preset value.

8. The picture ranking method of claim 1 , further comprising:
detecting second-type pictures from the user's pictures according to a facial recognition model, wherein the second-type pictures exclude a human face; and
ranking the second-type pictures according to a preset rule.

9. The picture ranking method of claim 8 , wherein detecting the second-type pictures according to the facial recognition model comprises:
extracting feature values of the user's pictures;
performing a matching between the feature values of the user's pictures and the facial recognition model; and
detecting the second-type pictures based on the second-type pictures that are less than a first preset value.

10. The picture ranking method of claim 1 , wherein before ranking the first-type pictures according to the social relations, the method further comprises:
acquiring interaction statistics between each of the contacts and the user; and
acquiring the social relations between the contacts associated with each of the first-type pictures and the user according to the interaction statistics.

11. A picture ranking device, comprising:
at least one processor; and
a non-transitory computer-readable storage medium coupled to the at least one processor and storing programming instructions, the programming instructions causing the at least one processor to be configured to:
acquire a user's pictures;
detect first-type pictures from the user's pictures, wherein the first-type pictures comprise one or more human faces;
determine contacts associated with each of the first-type pictures according to the one or more human faces in the first-type pictures;
determine social relations between the contacts associated with each of the first-type pictures and the user based on an identity feature of each of the contacts, wherein the identity feature of a contact describes a type of relationship between the user and each of the contacts;
rank the first-type pictures according to the social relations; and
present the first-type pictures after ranking the first-type pictures.

12. The picture ranking device of claim 11 , wherein the social relations between the contacts associated with each of the first-type pictures and the user are acquired according to a social relation model.

13. The picture ranking device of claim 12 , wherein the programming instructions further cause the at least one processor to be configured to
acquire information about the contacts associated with each of the first-type pictures using the social relation model, wherein the information about the contacts associated with each of the first-type pictures comprises the identity feature of each of the contacts.

14. The picture ranking device of claim 12 , wherein the programming instructions further cause the at least one processor to be configured to train the social relation model using a stored profile picture of the contacts and information about the contacts before the social relations are determined.

15. The picture ranking device of claim 14 , wherein the programming instructions further cause the at least one processor to be configured to:
extract profile picture data of the contacts; and
annotate the profile picture data using information about the contacts to obtain the social relation model, wherein the information about the contacts comprises the identity feature of each of the contacts.

16. The picture ranking device of claim 11 , wherein the first-type pictures are detected according to a facial recognition model.

17. The picture ranking device of claim 16 , wherein the programming instructions further cause the at least one processor to be configured to:
extract feature values of the user's pictures;
perform a matching between the feature values of the user's pictures and the facial recognition model; and
detect the first-type pictures based on the first-type pictures that are higher than a first preset value.

18. The picture ranking device of claim 11 , wherein the programming instructions further cause the at least one processor to be configured to:
detect second-type pictures from the user's pictures according to a facial recognition model, wherein the second-type pictures exclude a human face; and
rank the second-type pictures according to a preset rule.

19. The picture ranking device of claim 18 , wherein the programming instructions further cause the at least one processor to be configured to:
extract feature values of the user's pictures;
perform a matching between the feature values of the user's pictures and the facial recognition model; and
detect the second-type pictures based on a the second-type pictures that are less than a first preset value.

Assignees

• Huawei Tech Co Ltd

Inventors

• Jiang Zhuolin
• Yang Qiang

Classifications

• G06V40/172

  Classification, e.g. identification · CPC title

• G06N20/10 Primary

  using kernel methods, e.g. support vector machines [SVM] · CPC title

• G06V40/179

  metadata assisted face recognition · CPC title

• G06V20/30

  in albums, collections or shared content, e.g. social network photos or video · CPC title

• G06V40/168

  Feature extraction; Face representation · CPC title