WO2020051075A1

WO2020051075A1 - User identity determining method, apparatus, and electronic device

Info

Publication number: WO2020051075A1
Application number: PCT/US2019/048981
Authority: WO
Inventors: Dandan Zheng; Liang Li; Wei Xu
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba Group Holding Ltd
Priority date: 2018-09-04
Filing date: 2019-08-30
Publication date: 2020-03-12
Anticipated expiration: 2021-03-04
Also published as: WO2020051075A8; US11244199B2; US20200167598A1; TWI738011B; CN109344722B; CN111523569A; CN109344722A; US10997460B2; TW202011251A; US20200074219A1; CN111523569B

Abstract

The present invention discloses a user identity determining: method, apparatus and electronic device. The method includes: acquiring; target multidimensional feature information of a target user, wherein the target multidimensional feature information includes at least two types of feature information m biometric feature information and/or non-biometric feature information; comparing the target multidimensional feature information with multidimensional feature information of a plurality of designated users, respectively, to obtain a comparison result; and determining an identity of the target user based on the comparison result.

Description

USER IBENTITYBETERMlNIhiG METHOD, APPARATUS, AMD EEICTRONIC

DEVICE

Cross-Reference to Related Applications

{øøø 1 } This inierhatiimsl application is based upon and claims priority to Chinese Patent 5 Application No. 20181102502:2.8, filed on September 4, 2018, theenti® content of ail of which is incorporated herein l>y reference:,

Technical Field

10002} The present application relates to the field of computer technologies, and more particularly to a user identity determining method, apparatus and electronic device.

0 Technical Background

{0003] Biometric technologies are Widely rised in the identification of user identities because they are more convenient and secure. Biometric technologies are technologies that integrate computers with optical sensors, acoustic sensors, biosensors, the principles of biostafisties and other hi-tech Means to identify a personal identity using inherent physiological5 characteristics (such as fingerprint, thee, iris, or sound) of the human body.

0000} However, i n s om e eases,_: existing; solutions for usinghiomeirie features to identify user identities have a misidenfification problem for example, when face recognition is used to identify identities of twins, misidenlification often occurs. Also tor example, when voice recognition is used to identify two users having similar voices, misidenfificafioo may also occur ' > Sion m ary of the ! nvention

|000sf Embodiments: of the present application provide a user identity determining method, apparatus and electronic device to reduce user identity misidentification

10 00} To solve the echnical problems, the embodimen ts of the present application are Implemented as follows: A first aspect provides a user identity determining ethod, comprising:5 acquirin taigel multidimensional feature information of a target user, wherein the target multidimensional feature ihfortnatio comprises at least two types of feature information in biometric feature information and/or non-bferaetrie feature information; comparing fee target n itidimensionai feature information wife meliidimensionat feature info iatie» of a plurality of designated users, respectively, to obtain a comparison resul t; and determining an identi ty of 5 the target user based on the comparison result

1 0071 A secon aspect provides a: user identity determining apparatus, comprising; an acquiring module, configured to acquire target multidimensional feature information: of a target user, wherein the target nndtidimeBsienal feature infotn tion comprises at least two types of feature information in biometric feature information and/or non-biometric feature information; .10 a comparing module, configured to compare fee target multidimensional feature information with multidimensional feature information of a plurality of designated users, respectively; to obtain a comparison result: an &dotPrminiiigmodule, configured to determine an identity of the target user based on the comparison result.

{0908j A thir aspect provides an electronic device, comprising: a processor; and a 15 memory configured: to store computer executable instructions, wherein execution of the execniabfe msirtrcrions causes fee processor fo perform the following operations: acquiring target multidimensional feature information of a target user, wherei fee target nnfeidimensional featnre information comprises at least two types of feature information n biometric feature ioforinatjon and/or nou-biometric feature information; comparing the target multidimensional -o feature formalion wife nuriridimensioiml feature information of a plurality of designate users, respectively; to obtain a comparison result; aud determining an identity of the target user based on fee comparison result.

fOOOPj A fourth aspect provides a compute? readable storage medium, storing one o more programs, wherein fee one or more programs, when executed by an electronic device 25 comprising a plurality of applications, cause fee electronic device to perform fee following opeMions;; acquiring target multidimensional feature information of a target user, wherein the large! multidimensional featdre information comprises at least two types oi^' feature information in biometric feature information and/or non-biometrie feature information; comparing the target multidimensional feature information with multidimensional feature information of a plurality 5 of designated users, respectivel , to obtain a comparison resuit; and determining an identity of the target user based on the comparison result,

|0010| As can be seen from the technical solutions provided by the foregoing embodiments of the presen application, the; solutions provided by the embodiments of the present application at least have one of the following technical effects: because the identity of 10 the target user ts detenn ed based on the multi d imenskmal feature information of the target user, not based on single feature information such as face or voice, user identity isidentificaiion Can be reduced, thereby improving the reliability of the result of user identity determination.

Brief Description of the Drawings

jOtll l j The accompanying drawings described herein ar provide for further 15 comprehension of the present application, and constitu te a part of the present application. The schematic embodiments of the present application an the description: thereof are; used to illustrate the present application, but do not constitute any improper limitation to the present application. In the accompanying; drawings:

1QQI2| FIG. 3 is a: schematic flowchart of a user identity determining method according _ < * to an embodiment of this specification,

10013] FIG. 2 is a schematic diagram of building first database according to an embodiment of this specification.

|0014| FIG. 3 is a schematic di gram of face feature information at multiple poses according to an embodiment of tills specification.

L 01S| FIG. 4 si a schematic diagram of gait feature info ation according to an embodiment of this specification.

10O16J FIG. 5 is a detailed schematic flo wchart of step 104 in the embodiment shown in

FIG. 1.

5 P0J 7J FIG. 6 is a schematic diagram of an Index feature comparison process according to an embodiment of this specification,

p0l8] FIG. 7 is a schematic diagram of another index feature comparison process according to an embodiment of this specifica tion,

(0019} FIG. 8 is another detailed schematic flowchart of step 104 in the embodiment0 shown in FIG. 1.

P&20J FIG. 9 is a schematic structural diagram of an electronic device according to an embodiment of this specification.

pft21J FIG. 10 is a schematic structural diagram of a user identity determining apparatus 1000 according to an embodiment of this specification_:.

5 p022j FIG, 11 is a detailed schematic structural diagram of a module 1002 in the embodiment shown in FIG. 10.

(00231 FIG. 12 i another detailed schematic structural diagram of the module 1002 in the embodiment shown in FIG 10,

©«tailed Description

<> p024J To make die objectives, technical solutions and advantages of the present application dearer, the technical solutions of the present application will he described clearly an completely below with reference to the specific embodiments and corresponding accompan ing drawings of the present: application. Apparently, the described embodiments are only some rather tha all embodiments of the present application. On the basis of the5 embodiments m the present application, all other embodiments acquired by those of ordinary skill in the art without creative efforts shall belong to the protection scope of tile present application

P025 j 1b reduce user identity misidentilfcstion, the embodiments of this pecification provide user idenms determining method and apparatus The method and apparatus can be 5 applied to a surveillance system requiring user identity identification. For example, nowadays cameras are deploye thtonghout business districts, and merchants hope to use an existing video surveillance network in a business district to identify its. membership consumer based on information acquired by a camera immediately when the membership consumer appears in the business district, and rapidly: search historical consumption data and consumption habits of the0 membership consumer, so as to better en s the membership consumer. For ease of description, the determining of the identity of a user entering a business district is used as an example below to describe the user identity determining method and a paratus pros ided by th e bodiments of this specification,

{0ft26J A user identity determining method provided by the embodiments of this5 specification is describe in derail below with reference to FIG, f to FIG, 8.

|002?1 As shown in FIG, 1, a user identit determining metho provided by the embodiments of this specification may include the following steps: At step 102, acquiring target multidimensional .feature information: of a target uses; wherein the target oiuitidimensionaJ feature: information includes at: least two types of feature information in biometric feature«* information and/or non-btomeiric feature information

|0O28j The target user may be a user whose identity needs to be determined, identified, or authenticated, for example, a consumer entering a business district (lor example, a: shopping mall)

|O029| That: the target_: multidimensional feature information includes at least two types5 of feature information in biometric feature information and/or uou-biomeiric feature information nay mean that the ta get multidimensional biometrie feature foftnMioft includes at least two types of feature information s the iometric feature: information; or, the tar et nMliidimensionai biometric feature information includes at feast two types of feature information in the «on-biometric feature information: op die target multidimensional biometric feature infommhOu includes at least two types of feature information in the biometric feature information and the non-biometric feature information,

|0030] Fo example, the biometric feature information may include: face feature information, body feature information, gait feature information, cloth feature information, age feature information, gender feature information, and the like. Optionally, to further .reduce user identity misidentification, the face feature information may include face feature information at multiple poses, for example^ f ce feature information of a frontal face o a 45-degree side f ce. In practical applications, face feature information at multiple poses may be acquired by using a plurality of cameras. Likewise,_: the body feature information may also include body feature information at multiple poses, for example, body lea lure information from the front or a 45- degree side. The pit feature information may include gait feature information in walking and running states

10031] The non-biometric feature information may include: user ID information, geographical location information time information, Wi-Fi address information, and the like. The user TP informatio may further inbktde information uniquely identifying: foe user identity, for example, one or more of a user mobile phone number, a user identity number, and user mobile phone media access control (MAC) information. The time information may be information about the time at which the iden tity of foe target user needs to be determined,

f0032] When foe target multidimensional feabneinibrmatio includes biometric feature information of the target user, at step 102, an image of the target user may be acq uired by us mg an image capture device, for example, by using a camera disposed in foe business district, and

0 then the acquired image is analyzed to ohtairrthe biometric feature rnionnaho of the target use?, such as face feature information, body feature information, gau feature information, oth feature information, age feature in&r atioiy and geader feature foriaa on.

{0Q331 When the target nndtiduBensional feature information includes non-biometric 5 feature information of the target user, at step 102, ID information, geographical location information, and the like of the: target user may be obtained in other manners. For example, the mobile phone MA.C information of the target user may be detected by using Wl~Fi in the business district;, and used as the ID information of the target user; the location information of the target user may be determined aceordlng to location information of the business district.

K) 100341 At step 11)4, comparing the targe! multidimensional feature information with ninhidimeositmal feature information f a inra h ty of designated users, respectively, to obtain a comparison result

|0035J The designated user may be a pre-stored user whose identity is known, for example , a membership consumer pre-stored in a user identity surveillance system of a business 15 district,

|0036| If should he understood that by comparing the acquired target mui t dimen i on al feature information of the target user with the multidimensional feature information of the plurality of designated users whose identities are known similarity values of the target user with respect to th plurality of designated users may be respectively detefnhnedgao the determined similarity values are used as the comparison result, so as to determine the identity of the target user at ste 106.

{0037J In an example, a weighted sum of similarity values corresponding to various feature information in foe multidimensional feature information may be used as a similarity value of the target user with respect to a designated user. For example, it is assumed that four pieces 2S of feature information A, B, C, and D of the target user are respectively compared with four pieces of feature information A, B, C, and D of the designated user, and four siMlarity values a, b, c, and d may be correspondingly Obtained, hi ibis case, a weighted sum of the similarity values a, b, c, add d may be used as rite similarity value of the target user with respect to the designated user, Weights corresponding to different feature information may be determined according to the degree of ««parlance of the feature Infbrmati o ,

{0038 j in practical, applications, a first database storin a, correspondence between user ID information of designa ted user and muhi dtmensiotial feature information of the designated user may be built in advance, so feat during comparison, fee target multi dimensionai feature information of the target user is compared with fee multidimensional feature information of the designated user in the first database built in advance, to obtain the comparison result.

}0039| The process of budding the first database Is briefly described below wife reference to F1U 2 to FIG, 4,

{0040] in an example, as shown in FIG, 2, it is assumed feat fee multidimensional feature informati n_^ of the: designated user snci tides biometric feature information, wherein tire biometric feature mformaiion specifically includes: face feature information, body feature information, gait feature information, cloth feature information, age feature information, an gender feature information. As can be seen from FIG. 3, the face feature information may include lace feature information For different fecial expression and different poses. As con be seen from FIG 4, fee gad feature information may include gait feature mfemiaiion in walking and running states p»4il Specifically as shown In FIG. 2. for a designated user 1 {corresponding to user

ID information represented by IDI), biometric feature inlcimalion sticb as face l face 2, body 1, body 2, gait 1, gad 2, age, and gender may be acquired from a surveillance video captured by a camera, and store in the first database. Likewise: for a, designated user 2 (corresponding to user ID information represented by ID2). biometric feature information such as face 1, face 2,

b body 1 , body 2, gait 1 /gait 2, age, and ender may be acquired train a surveillance video captured by a camera, and stored in the l UNI database

P04 1 In addition, as shown in FIG. 2, because the biometric feature information of the user changes with time or seasons, the biometric feature information of die designated use 1 and 5 the designated user 2 at different time points (time point I and time point 2 in FIG 2) may be recorded and saved. Optionally the first database may further record: and save spatial dimensional features of the designated user, which specifically include feature information such face, body, and gait of the designated user captured by different cameras at different: angles, {06431 Moreover, as shown in FIG 2, a relationship between users may further he0 considered, and the relationship mas also be stored in the first database.

}0044| in another example, the non-biometrie ^"feature hvforrmition of the designated user and the ID information of the designated user may also be corresponding iy stored in the first database.

j0045j At step 106. deiermimng an identity of the targe user based on the comparison5 result

|0046j In th exa ple used in step 104, if the comparison result is the similarity values of the target user with respect to the plurality of designated users, at step 100, an identity of a designated user in die plurality of designated tisers that has the largest similarity value, which is greater than a preset threshold, with respect to the target user may be determined as the identity < * of the target user; correspondingly, ti the plurality of designated users do not include a designated user whose degree of similarity to the target user is greater than the preset threshold, the target user may be determined as a new user or a non-membership user. The preset threshold may be manually set. For example, the preset threshold may be 80%

f 00471 !p the user identity determining method provided by the embodiments of this5 specification, because the identity of the target user is etermined base on the multidimensional feature mforrasitba of die target user, not baser! on single feature information such as fece or voice. User identity rmsidentiffeatfen can be reduced, thereby improving the reliability Of the result of user identity determination

]0 8j Because the targe user generally needs to be compared with a large number of 5 designated users (that i . the first database stores a large number of designate users), directly comparing the target multidimensional_: feature : information of the target riser with the multidimensional feature Information of the plurality of designated users involves comparison of a large amount of data, leading to low comparison efficienc and accordingly low efficiency of determining the identify of the target risen

0 10049 j To increase the efficiency of determining the ide iiy of ihe targei user, optional iy in another embodiment as shown in FIG. S, step 104 may includ the following substeps: fOOSO] Substep 5< >2. determining a plurality of index features of the target user based on the target muhidimensioi l feature information,

{0 51] It should: be understood that the index feature is a feature uniquely identifying5 original feature information (for example, the feature information acquired in step 102) and having a data amount less than that of the original feature information.

1005 ] The plurality of index features may include: a single inde fea ure for the same type of feature information, a composite index feature for at feast iwo types of feature information oecnriing at the same time, or a composite index feature for feature: information of < > different users occurring at the same time,

10053] The single index feature may be an index feature determined by on type of feature information, The composite index feature may be an index feature determined by two or mom types of feature information.

10054] For example, die single index feature for the same type of feature information5 may be: a geographical location index feature for the geographical location information, a face index feature for lace feature infermaiiou at a particular angle, or the life. For example, the composite index fea lure for at least two types of feature information occurring at die same time may be: a composite index feature formed by an obvious body feature and an obvious doth feature. For example, tire composite: index feature for feature information of different users 5 occurring at tbe same time may be: a composite index feature formed by feces of a couple shopping at fee: same time in a shopping mall

fOOSsj The process of determining a plurality of index features of the target user is described below by using examples.

jO0S6| in a first example, assuming that fee target multidimensional infermMion of fee0 target user includes the geographical location information of fee target user, substep 502 may include: determining a level 1 geographical location index feature of the target user based M the geographical location information of the target user; or, determining a level I geographical location index feature an a level 2 geographical location index feature of the target user based on the geographical loesfinn mfermafioBof the terget user; or. determining a level 1 geegrapMca!5 location index feature, a level 2 geographical location index feature, and a level 3 geographical location index feature o fee tarpt user based: on fee geographical location information of die target user. The level 3 geographical location index feature is a subindex of the level 2 geographical location index, and the level 2 geographical location index is a subindex of the level 1 geographical location index.

<> poSTf For example, an area wife a fixed flow of people, such as a shopping mall, mansion, office building, or a community, may be determined: as the level i geographical loeatios index feature; a floor or store in fee shopping mail, a floor or room in the mansion, or a unit or floor of the community is determined as fee level 2 geographical location index· feature; a location of a cam r _: in the stare in the shopping ail is determined as the level 3 geographical5 location index feature. ! b$$! la a second example assuming that ihe target multidimensional feature information includes the hfometoc feature information of he target user; for exa ple, face feature mfonaatida, body feature information, and doth feature information, suhstep 502 .may include; performing aPrinetpa! Component Analysis (PC A) on the biometric feature information 5 of the target user to obtain redueed-dhnensionaiity features of the biometric feature information; buefcetizing_: the reduced-diffiensiorsality features in a plurality of buckets representing the biometric feature information; and determining IDs of the plurality of buckets as tire plurality of index features of the target user. For example, a gait: bucket index feature, a clotli bucket index feature, a gender bucket index feature, a face bucket index feature, an age bucket index feature,0 a body bucket index feature, and the like are obtained. A bucket ID a be a serial number such as 1, 2, 3, or 4,

[00S9| In addition, assuming that the target multidimensional feature information includes the ID information and the time information of the target user, a time index feature (or referred to as timestamp) and an IB index feature ma further be determine based on substep5 502, Based on th target multidimensional feature information, a relalion-feee-eharaeter composite inde feature, a feee-body composite index feature, and the like are determined, jO fi] Substep 51)4, comparing the plurality of index features with a plurality of index features: of the plurality of designated users, respectively, and determining_: a pforulity of first users from the plurali ty of des ignated users.

<> 0i½ 11 Similar to the embodiment shown in RG. 1 , in practical applications, a second database storing a correspondence between user ID information of a designated user an a plurality of index features of the designated user nmy also be built in advance, so that during comparison, the plurality of inde features of the target user may he compared with the plurality of index features of the designated user in the second database, to obtain a piurality of first users,5 The plurality of index features of the designated user is determined^" in a maimer similar to that in which the plurality of index features of the target user is determined, and the details will not be repeatedly described here,

£8062] For example, finally determined index features stored in the second databas may include; a gait bucket index feature,* a cloth Bucket index feature:, a gender bucket index feature, 5 a face Bucket index feature, an age Bucket index feature, a body bucket inde feature, a time index feature: (or referred to as timestamp), an ID index feature, a relational-face-character composite index feature, a face-body composite index feature, and the like.

10063] It should Be noted: that in the embodiments provided; by this specification, the -first database and the second database may fee the same database or may fee different databases:0 fO064] As an example, in subsiep 504, the plurality of index features of the target user may fee respectively compared with the index features of the plurality of designated users, to determine, from the plurality of designated risers, users having one or more of the plurality of index features matching corresponding index features of the tarpt user as the plurality of first users. It should be understood that the index features matching each other means that the index5 features are the same or similar for example, assuming that the plurality of index features of the target user includes the geographical location index feature and the face inde feature, plural h \ of users whose geographical location index feature and face index feature are the same as the geographical location index feature and the face index feature of the target user nun h found from the plurality of designated users as the plurality of first users.

< > pOdSI As another example, in substep 504, the plural u> of mdex features of the target user may be sorted according to a certain rule, tor example, sorted at random or according to the degree of accuracy of user identity identification; afterward, a plurality of users matching the index feature ranke the first are determined from the plurality of designated users and then, front the plurality of users, a plurality of users matching the index feature ranked the second are5 determined, and so on. in this way; a plurality o f users matching an index feature is determined Item a determining result obtained according to the matching of an index: feature ranked before that index feature. Finally, a plurality of users matching the index feature ranked the last i> used as the plurality of first users. The process of d etermi rung the plurality of first users in the manner shown In the example is described below with reference to FIG. 6 and FIG. 7.

|00661 As shown in FIG. 6, first, multidimensional feature information 601 including fe.ee 1, face 2, and face 3 of the target user is acquired. Specifically, three pieces of feature information“face I, face 2, and face 3” of the target user may be acquired by using a plurality of cameras at different: angles at a time at which th identity of the target user needs to be determined. Second, based on the acquired multidimensional feature information of the target user, a plurality of index features of the target user are determined Then, the plurality of index features of the target user are compared with corresponding index features (for example, composite index feature-relational face feature, d single index feature- face feature) of a plurality of designated users stored in a secon database 602, to determine x first users. Finally, in a manner shown: in substep SOP below, foe three pieces of feature information“fece I , fece 2, and, fe.ce 3"' aroused as prticnkr feature information of the target user and compared wit “particular feature information of foe x: first users” 603, an similarity values of the target user with respect to the x first users are determined.

{0 7| / s shown in FIG. 7, first, multidimensional feature information 601 including fees 1 , fece 2, and fece 3 of the target user is acquired. Specifically, three pieces of feature informat ion“face I , lace 2, and face 3” of the targe t user may he acquire b using a plurality of cameras at different angles at a time at which the identity of the target user needs to he determined. Second, based on foe acquire multidimensional feature information of foe target user, a plurality of index features of the target user are determined. Then, the plurality of index features of the target user are compared with corresponding index features (for example, composite index feature-relational face feature, or single index feature-feee feature and Wi-Fi address 604} of a plurali ty of designated users Stored in a second database 602, to determine x first users Finally, in a manner shown in sbibstep 506 below, die three pieces of feature information“face 1, lace 2, and lace 3” are used as particular feature information of the target user and compared: with '‘particular feature information of the x first users” 603, and similarity 5 values of the target user with respect to the x first users are determined.

{006$ j Subsie 506, comparing particular feature information In the target multidimensional feature Information with particular feature Information of the plurality of Fust users, respectively an determining similarity values of the target user with respect to the plurality of first users.

0 f0069| in subsiep 506, based on the plurality of first users determined in substep 504, particular feature information is further compared, and the similarity values of the target user with respect to the plurality of first users are respective!} determined,

|007q| The particular feature information may he biometric feature information of the target user, for example, face feature information, body feature information, cloth feature 5 informaiion, or the like,

f00†J{ Snbstep 508, determining the similarity values of the target user with respect to the plurality of first users as the compari on result.

{0;072| Based ptr this, step IQfi may include: determining an identity of a: use corresponding to the largest similarity value which is greater than a preset threshold to the <> comparison result as the identity of the target user. That is, in the plurality of first users, an identity of a user that has the largest similarity value, which is greater_: than the preset threshold, with respect to the target user is determined as tire identity of the target user,

|O073{ It should be understood that in the embodiment shown in FKl 5, because Index features having a small amount of data are compared first, a plurality of candidate users i that is,5 a plurality of first users) that may be .similar to the target user are determined from the plurality of designated users so" as to reduce the range to be compared, and then iri foe small range, the identity of tile target user is determined through comparison of the particular feature in!brraaiion, the efficiency of determining the identity of the target user can be impro e ;

{0Q741 lb Improve die efficiency bf determining the Identity of the target user, optionall y 5 In still another embodiment as shown in FIG, 8, step KM may include the following suhsteps;

{0075 j Snhsiep 802 determining a plurality of index feature of the target user based on the target multi imensional feature Mormatian.

100761 Substep S04, comparing the plurality of index features with a plurality of index features of the plurality of designated users, respectively, and determining plurality of first0 users from the plurality of designated users,

J0077{ Specific implementations of substep 802 and substep 804 are correspondingly the same as those of subste 502 and subsrep 504 in the embodiment shown in FIG, 5, so the details will not: be repeated herein.

{0078 j Substep 806, selecting a plurality of second users frenithe plnraht₎ of first users 5 based on historical behavior ata of the plurality of first users,

{0070 j In an example, first the plurality of firsf sejs ma be sorte based on historical behavior data of foe plurality of first users, an then a plurality of second users whose rankings satisfy a preset condition is selected from die plural uy of first users.

{00801 When foe deteiiuining of an, identi ty of a user entering busines district is used <> as an example, foe historical behavior data may be a quantity of historical consumptions, a quantity of historical entr times, or a historical consumption amount; correspondingly, when the. plurality of first users Is sorted, the first users may be sorted in a descending order of the quantity of historical consumptions, the quantity of historical entry times, or the historical consumption amount. £0081] lu another example, the plurality of first users may be sorted based on a preset sorting: model, and then a plurality of second users whose rankings satisfy a preset condition is selecte feorn the plurality of first users. The preset sorting model may be a sorting model in the rsor an.

5 {W82J The step of selecting: a plurality of second users whose rankings satisfy a preset condition from the plurality of first users may include; selecting the top-ranked several users from tire plurality of first users as the plurality of second users.

{0083 j Substep 80S. comparing particular feature information in the target multidimensional feature information with particular feature information of the plurality of0 second users, respectively, and determining similarity values of the target user with respect to the plurality of second users.

ίO084| Siibstep 8Ϊ0, determining the similarity values of the target user with respect to the plurality of second users a the comparison result.

pOSSj Specific implementations of substep >8 and substep 810 are eocreSpondiugly the5 same as; those of siibstep 506 and. substep 508 in the embodiment shown in FIG. 5, s the details will not be repeated herein,

{0086] Based on the embodiment shown in FIG. 8, step 1:06 ma include: determining an identity of a user corresponding to the largest si ilarity value which Is greater than a preset threshold in the comparison result as the Identity of the target user. That is. In the pluralit of«> second users, an identity of a user that has the largest similarity value, which is greater than the preset threshold, with respect to the target user is determined as the identity of the target user

{0087J Compared with the embodiment shown in FIG, 5, in the embodiment shown in FIG. 8, not only index features: having a small amount of data are compared first and a plural h of candidate users (that Is, a plurality of first users) that may be similar to fee target user are5 determine from fee plurality of designated users so as to reduce the range to be compared, but also a plurality of second users is fur i er selected from the plurality of first users to further reduce tlic range to Be compared; then in the further reduced; range, the identity of die target user is determine throug comparison of the particular feature information. Therefore, the: efficiency of determining the identity of the target user ean be further improved.

fO08¾l The above is a description of a user identity determining method provided by this specification. An electronic device provided by this specification is described below.

|0089J FIG. 9 is a schematic structural diagram of an electronic device according to an embodiment of this specification. Referring to FIG; 9, at the hardware level, the; electron c device includes a processor; and optionally can further include an internal bus, a network interface, and a memory. The memor may include an internal memory, such as a high-speed Random Access Memory (RAMf and may farther include a non-volatile memory, such as at least one magnetic disk memory. The electronic device can further include other hardware required by services. ftMi !j The processor; the network interface, and the memory may be connected to each the through the internal bus. The internal bus ma he an Industry Standard Amhiiechrre (ISAI bus, a Peripheral Component interconnect (PCD bus, an Extended Industry Standard Architecture (EISA) bus, or the like, The bus may include on address bus, a data fens, a control bus, or the like. For ease of illustration, only one bidirectional arrow· is used i FIG. 9 to represent the bus, but it does not represent that there Is only one bus or only one type of buses.

109911 The memory is configure to store a program. Specifically, the program may incltde program codes, and the program codes include computer operation instructions. The memory can Include an internal memory and a non-volatile memory, and provide: iosiruetioiis and data for the proces sor.

100 21 The processor can rea a corresponding computer program from the non-volatile memory into the internal memory and then run the program, to form a user identity determinin

IS apparatus at the logical le vel. The processor can execute the program store in tile memory, and is specifically configured to perform the fol towing operations;

P0 3J ac uiring target multidimensional feature information of a target user, wherein the target multidimensional feature information includes at least two types O feature information 5 In biometric feature information and/or non-biomeirie feature information;

[00®4| compering die target multidimensional feature information with multidimensional feature intormation of a pl urality of designated users, respectively, to obtain a comparison result; an

1 953 determining an iden tity of the target user based on t he comparison result 10 [00961 I he user uienlity determining method disclosed in the embodiment shown in FIG.

1 of this specification can be applied to the processor or implernehled by the processor. The processor may he an integrated circuit chip having a signal processing capability. In an iaplememation process, the step of the foregoing method can he co piete by hardware integrate logic, circuits in: e processor or software-form instructions The processor may be a 15 general-purpose processor including a Central Processing Uni (CPU), a Network Processor (hiP), or the like, and may also bo a Digital Signal Processor (DSP) an Application Specific integrated, Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or another- programmable logic device, a discrete gate, a transistor logic device, or a discrete hardware component: The processor can implement or execute various methods, steps and logic block diagrams disclose i one or -«> more embodiments of this specification. The genera! 'purpose processor may be a micro processor, or any regular processor or the like. Steps of the method disclosed in, combination with one or more embodiments of this specification can be directly embodied as being execute by a hardware decoding processor or executed by a combination of hardware in the decoding processor and software modules. The: software modules can, be located to a mature, storage 25 medium in the art, such as a random access memory, a flash memory, a read-paly memory, a programmable read-only memory, a electrically erasable programmable read-only memory, or a register. The storage medium i located m the memory. The processor reads information in die memory, and completes the steps of the method in co bination with the hardware of the processor

5 1009?! The electronic device may farther perform the user identity determining method of BO 1 , and the details will not be repeated herein

|0098] In addition to the software implementation, the electronic device of this specification does not exclude other implementations,, suc as a logic device, or a combination of hardware and software. In other words, entities executing the following processing procedure0 are not limited to logic units, but may also be hardware or logic devices

}0099| The enibodintents of tins specification further provide a computer readable storage medians. The computer -r adable storage medium stores one or more programs. The one or more programs include m iustmction . The stmction, when executed by a portable electronic device including a plurality of applications., causes the portable electronic device to perform the 5 method of the embodiment shown ½ FIG 1 to specifically perform the foil owin operations:

{Of 0(1] acquiring targe multidimensional feature information of a target user, wherein the target multidimensional feature information includes at least two types of feature information m biometric feature information and/or npn-biomeiric feature Information:;

10181] comparing the target multidimensional feature information with <* multidimensional feature infomiation of a plurality of designated users, respectively, to obtain a comparison result; and

|OI02j determining an identity of the target user based on the comparison result

]0183] A user identi ty determining apparatus provided_: by this speci fication is described below. |01¾4| FID. 10 is a sdiematk structural diagram of a user identity detemmlfog apparatus iH! i according to tills specification, Referring to FIG, 10, in a sofsvare snipienieniaiion, foe user ideniitv determining apparatus 1000 may include an acquiring module 1001, a comparing module HK!2, snd a determining Module 1003,

1010S1 The acquiring module 1001 is configured to acquire target multidimensional feature information of a target user, wherein the target multidimensional feature information includes at least two types of feature information in, biometric feature information and/or non- hionieir . feature information.

j01061 The biometric feature information includes: lace feature information, body feature information, gait feature information., cloth feature information, age feature infonnation, an gender feature information; and tile non-biometric feature information includes user ID information, geographical location information, and time information, wherein foe user ID information includes one or more of a user mobile phone number, a user identity number, and user mobile phone MAC information,

|0ii>7l The comparing module 100 is configured to compare foe targetmultidimensional feature information with multidimensional feature information of a plurality of designated users, respectively, to obtain a comparison result.

f i 8} As an eyampfe, foe comparing module 1002 may he specifically configu e to: compare th target multidimensional feature information with foe multidimensional feafore information of ihe plurality of des ignated users, respectively, to Obtain similari ty values of the target user with respect ip the plurality of designated ttsers: and determine the similarity values of foe target user with respect to the plurality of designated users as the compariso result.

1010.9} The determining module 1003 is con figured: to determi ne an identity of the target user based on fee comparison result

2 ! |il 101 la the user identity determining apparatus provided by the embodiments of this specification, because the identity of file target user is determined based on the multidimensional feature information of the target user, not based on single feature information such as face or voice, user identity misidentification can he reduced, thereby improving the reliability of the 5 result of user identity determination.

{0111} Optionally, in aaofeer eniboditneni as shown in FIG. 11 , tile comparing module 1002 may specifically include; a first index dete mining submodule 1101, a first user determining: suhtuodale: 1102, a first: similarit value determining submodule 1,103, and a first resu l t determining submodul e 1104.

0 {0112} The fu st index determining submodule 1 101 is configured to determine a plurality of Index features of the target user based on the target multidimensional feature information, p:!:l3| The plurality of index features include: a single index feature for the same type of feature information, a composite index feature for at least two types of feature information occurring at the same time, or a composite index feature for feature information of different users 5 occurring at the same time,

{0114} A an example, if the target multidimensional feature information includes geographical location information of the target user, the determining a plurali ty of index features of the target user based on the target nndiidimensionai feature Information indudes: determining a level f geographical location index feature of the target user based on the geographical location <* information of the targe! user; or, determining a level 1 geograplncal location index feature and a level 2 geographical loca tion index feature of the target user based on the geographical location information of fee target user; or, determining a level 1 geographical location index feature, a level 2 geographical location index feature, and a level 3 geographical location index feature of the target user based on the geographical location information of fee target user. The level 3 geographical location index feature is a snhitidex of the level 2 geographical location index, and fee level 2 geographical locaiion index is a subindex of fee level 1 geographical location index PI I SJ As another example, if the target multidimensional feature information includes biometric feature inldm lidn of the target user fee determining a: plurality of index features of 5 the target user based on the target multidimensional feature information includes: performing a principal component analysts tPCA) on the biometric feature information of the target user to obtain rsducedfeimensionality features of the biometric feature information; bucketixfeg fee reduced-dimensiouaiity features in a plurality of buckets representing the biometric feature information; and detonluing IDs of fee plurality of buckets as fee_: plurality of index features of0 the target user.

}0116| The first user determining ubmodule 1102 is configured to compare the plurali t of index features ufe a plurality of index features of the plurality of designated users, respectively, an determine a plurality of first users from fee plurality of designated users. piiTj As an example, the first use determining submodule ! 102 may compare fee5 plurality of index features of the target user with the index features of the plurality of designated users, respect! vely, to determine, from the pi araiiiy of designate users, users having one or more of the plurality of index .features matching corresponding index features of fee target user as fee plurality of first users it should he understood feat the index: features matching each ether means feat fee index features are fee same or similar.

«> pi!$J As another example, the first user defermimng submodule 1102 may first sort fee plurality of index features of fee target user according to a certain rule, for example, at random or according to the degree of accuracy of user identity identification; afterward, determine a plurality of users matching fee index feature ranked the first from he pluralit of designated users, and then determine a plurality of users matching the index feature ranked fee secon H orn5 fee plurality of users, and so on. In this way, a plurality of users matching one index feature: is determined from a determining result obtained according to the matching of an index. feature ranked before the index feature. Finally, a plurality of users matcltitig the index feature ranked the last is used as the plurality of first users.

{01191 The first similarity value determiuhtg; submodule 1103 is configured to compare particular feature information in the target multi dimensional feature information with particular feature information of the plurality of first users, respectively, ard determine similarity values of the target user with respect to foe plurality offirst users.

1 120! As an example, based on the plurality of first users deie.rmiued by th first user determining submodule 1 102. the first similarity value determining submodule 1103 further compares panicular feature information and determines the similarity values of the target user with respect to the pl urali ty of first users, respectively,

{0:!2I| The panicular feature information may be biometric feature information of the target user, for example, foes feature information, body feature information, cloth feature information, or the like.

|0l22j Tbs first result determining submodule 1104 is configured to determine the similarity values of foe target user with respect to: the plurality of first users as fee comparison result,

{0123{ Based on this, the determining module 1003 may be specifically configured to deterroine an identity of a user corresponding to he largest simi larity value which is greater than a preset threshold in the comparison result as fee identity of the target user. That is, in the plurality of first users, an Identity of a user that has the largest similarity value, which is greater than the preset threshold_* with respect to fee target user is determined as the identi ty of the target user,

{0124{ It should be understood that in the embodiment shown in FIG. 11 , because index features having a small amount of data are compared first, a plurality of candidate users t that is, a plurality of first users) that may be similar to the target user are determined fro the plurality of designated users so" as to reduce the range to be compared, and then in the small range, the identity of the target user is determined through comparison of the particular feature information, the efficiency of determining the identity of the target user can be improved;

{0I25| Optionally in still another erriboditnehi as s own in FIG, 12:, the comparing 5 module 1002 may specifically include: a second index determining submodule 1201, a second user determining submodule 1202, a third user determining submodule 1203, a second similarity value determining submodule 1204, and a second result determining submodule 1205,

10126] The second index determining submodule 1201 Is configured to determine a plurality of index features of the target user based on the target multidimensional feature0 information

}0127{ The second user determining submodule 1202 is configured to compare the plurality of index features with a plurality of index features of the plurality of designated users, respectively, an determine a plurali ty of first users from the p lural ity of designated users,

{0128] Specific implementa tions of the second index determining submuduie 1201 an5 the second user determining submodule 1202 are correspondingly the same as those of the first index determining submodule i 101 and the fust user determining submodule 1102 in the embodiment show i FIG. 11 , so the details will not be repeated herein

{0129{ The third user determining submodule 1203 is configured to select a plurality of second users from the pl ural tty of first users based oh historical behavior da ta of the plurali ty of <> first users,

|0138j in an example, the thir user determining_: submodule 1203 may first sori tire plurality of first users based on historical behavior data of the plurality of first users, and then a plurality of second, users whose rankings satisf a preset condition are selected from the: pluralit of first users. 10131] In another example, the third user determining submodule 1203 may sort the plurality of first users based on a preset sorting model, and then select a plurality of second users whose rankings satisfy a preset condition from the plurality of first users. The preset sorting mode! may he a sorting model in the prior art

5 101321 The third user determining submodule 1203 selecting a plurality of second users whose rankings satisfy a preset condition from the plurality of first users may include; selecting; the top-ranked several users from the plurality of first users as the plurality of second users.

10133] The second similarity value determining submodule 1204 is configured to compare particular feature information in the target multidimensional feature information with0 particular feature information of the plurality of second users, respectively. and determine similarity values of the target user with respect to the plurality of second users.

10134] The second result determining submodule 1-205 is configured to determine the similarity values of the target user with respect to the plurality of second users as the comparison result.

5 1 135] Specific implementations of the second simi lari ty value determining submodule

1204 and the second result determinin submodule 1205 are correspondingly similar to thos of the first similarity value determining submodule 11.03 and the first result determining submodule 1104 in the embodiment shown in FtGi 1 1 , so the details w ill not be repeated herein.

10136] Based on the embodiment shown In F1G, 12, the determining module 1003 may«> be specifically configured to determine an identity of user corresponding to the largest Similarity value which is greater than a preset thres old in _: the comparison result as the identity of the target user. That is, in the plurality of second users, an identity of a user that has tire largest similarity value, which is greater than: the preset threshold, with respect to the target user is determined as the identity of the target user. P137 Compared with the embodiment shown in FIG. 1 1, in the embodiment shown in FIG. 12:, not nly index features haying a small amount of data are compared first and a plurality of candidate users {that is, a plurality of first users) that may be similar to the target user are determine from the plurality of designated: users: So as to reduce the range to he compared, but 5 also a plurality of second risers is further selected from the plurality of first users to further reduce tire range to be compared; then in the further reduced range, the identity of the target user Is determined through comparison of the particular featur information. Therefore, the efficiency of determining the identity of the target: user can be further improved.

j0l38| It should fee noted that the user identity determining apparatus 1000 can0 implement the metho of the method embodiment of FIG. 1 , the user identity determining method of the embodiment shown in FIG I ban be referred id for details, and the details will not be repeated herein,

10139) The above descriptions are merely preferred embodiments of this specification, and are not intended to limit th protection scope of this Specification Any modifications,5 equivalent replacements or improvements mad without departing from the spirit an principle of one nr more embodiments of this specification shall he encompassed in the protection scope alone or more embodiments of this, speeification.

|014f)| The system, apparatus, _.modules or units illustrated In the foregoing embodiments specifically an be implemented fey a computer chip or an entity, oHmpIememed by a product < * ha\mg a particular function A lypicai lmptaentation device is a computer. Specifically, the computer forexampfc can fee s personal computer, a lapto computer; a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an e-mail device, a game console, a tablet: computer, a wearable device, or a combination of any of these devices. P141J A computer readable medium includes noh-volahfe and volatile media as well as mosahte and non-movable media, and can implement information storage by means of any method or technology information may be a computer readable instruction, a data structure, and a modnle of a program Or Other data. For example a storage medium Of a computer includes, but 5 rs not limited to, a phase change memory (PRAM), a static rando access memory (SRAM), a dynamic random access memor (DRAM), Other ty es of random access memories (RAMs), a read-only memory (ROM), an electrically erasable programmable read-only memory (EERROM), a flash memory pr other memory technologies, a compact disk read-only memory (CD-ROM), a digital versatile disc (DVD) or other optical storages, a cassette tape, a magnetic 10 tape magnetic disk storage or other magnetic storage devices, or any other non-transmission media, and can be use t share information accessible to the computing device. According to the definition herein, the computer readable medium does not include transitory media, such as modulated data signals and carriers,

|M42j It should be feather noted that the terms ^s include, “comprise” or any other 15 variations t hereof are intended to cover non-exclusive incl usion, so that a process, method, article or device including a series of elements not only includes the: elements, but also Includes other elements not expressly listed, or further Includes /inherent elements of the process, method, article or device. In the absence of more limitations, an; element defined by“ including a/an.,” does not exclude that the process, method, article or device including the element farther has other Z<> identical elements.

|0143) The embo iments in dm specification are described progressively. Identical or similar parts of the embodiments may be obtained with reference to each other. Each embodiment focuses on description of apart different from other embodiments. Especially, the system embodiment is basically similar to the metho embodiment, so it is described simply. 2S For relate parts, refer to the descriptions of the parts in the method embodiment.

Claims

1. A user identity determining method, comprising;

acquiring target nudtidi ensiona! feature information of a target user, wherein the target multidimensional feature information comprises at least two types of featur information: in 5 biometric feature information and/or non-blometric feature information;

comparing the: target multidimensional feature information with multidimensional feature information of a plurality of designated users, respecti vely; to obtain a compari.son result; and determining an identity of the target user based on the comparison result.

.10 2. The method according to claim 1 , wherein the comparing the target multidimensional feature inforitat i nwith multidimensional feature information of a plurality of designated users, respectix ely, to obtain a comparison result comprises:

comparing the target multidimensional feature information with the multidimensional feature information of the plurali ty of designated risers, respectively, to ohhrin similarity values 15 of the target user with respect to the plurality of designated users; and

determining the similarity values of the target user with respeet to the plurality of designated_: users as the comparison result.

3. The method Recording to claim 1, wherein the comparing the target mnltidimensionai -o feature iiiformation with multidimensional feature information of a plurality of designated users, respectively; to obtain a com rison result comprises:

determining a plurality of index features of the target user base on the target niultidimensfonal feature information;

comparing the plurality of index features with a.plurality of index features of the plurality 25 of designate users, respectively, and determining: a plurality of first users If our the plurality of designated users;

comparing particular featmn information in th target mulftdi etsionai feature i formation with particular feature information of the plurality of first users, respectively, and determining similarity values of fee target user with respect to fee plurality of first Users; an

5 determining fee similarity values of fee target user wife respect to fee plurality o f in si users as the comparison result.

4. The method accordin g to claim 1, wherein the comparing fee target multidimensional feature information with multidimensional feature information of a plurality of designated users,

10 respectively fo obtain a comparison result comprises;

determining a plurality of inde features of the target user based On fee target multidimeftsionai feature information;

comparing the plurality of index features with a plurality of index features of fee pluralit of designate users, respectively, and determining a plurality of first users from fee plurality of 15 designated users;

selecting a plurality of second users fro the plurality of first users based on historical behavior data of h plurality of first users;

comparing particular feature information in fee target multidimensional feature mferanario wife particular feature information o f fee plurality of second users, respectively, and determining -o si mi lari ty values of the target user with respect to the p I ural tty of second users; and

determining fee similarity values of the target user wife respect to the plurality of second users as the comparison result

5. The method according: to, claim 3 OG4,_> wherein fee plurality of index features comprise; 25 a single index feature lor fee same type of feature inlonuaiion, a composite index feature lor at least two types of feature information occurring at the same time, or a composite index feature &r feature information of different users occurring at the same time.

6. The method according fo claim 3 or 4, wherein

5 the biometric feature information comprises: face feature information, body feature information, gait feature information, cloth feature information, age feature information, and gender feature information; and

the non-biometric feature information comprises user identification (ID) information, geographical location information, and time information, wherein the user ID information0 comprises one or more of a user mobile phone number a user identity number, and user mobile phone media access control (MAC) information.

7 The method according to claim 6, wherein if the target multidimensional feature information comprises geographical location: information of foe target user, the determining a5 plurality of index fea res of the target user based on the target multidimensional feature informatio comprises:

determining a level I geographical location index feature of the target user base on the geographical location Information of the target user; or

determining: the level 1 geographical location index: feature and a level 2 geographical < * location index feature of fee target user based on he geographical location information of fee target user; or

determining the level 1 geographical location index feature, fee level 2 geographical location index feature, an a level 3 geographical location i ndex feature of foe target user based on the geographical location information of foe target user;

S wherein the level 3 geographical location index feature is a snbindex of the level 2 geographical location index, and the level geographical location index is a subindex Of the level 1 geographical location index.

I. The method according to claim 6, where® if the target nliiditnenstonal feature 5 information comprises biometric feature information of the; target user, the determining a phuaSity of index feainres of the target user based on the target muStidimensional feature information comprises;

performing a, principal component analysis; fPCA) on the biometric feature information of the target user to obtain reduced-dimensionality features of the biometric feature information; 0 buckeifelng the reduced-dimensionality feature in a plurality of buckets representing the biometric feature information; and

determining IDs of the plural tty of buckets as the plurality of index features of the target 5 9 ^'the method record ng to an one of claims 2 to 4, wherein the determining an identity of the target: user based on the comparison result comprises :

determining an identity of a user corresponding; to the largest similarity value whic is greater than a prese threshold in. the comparison result as the i dentity of the target user «* 10. A user identity determining apparatus* comprising:

an acquiring module, eonhgure to acquire target niultidinien ionai feature information of a. target user, wherein the target multidimensional feature information comprises at least two types of feature information in biometric feature: information and/or non-biometrie feature Information;

5 a comparing module configured to compare the target multidimensional feature information with multidimensional feature information of a plurality of esignated users, respectively, to obtain a compariso result; and

a determining module, configured to determine an identity of the target user based on tie comparison result

5

O An electronic_, device, comprising:

a processor ; and

a memory configured to store computer executahie instructions, wherein execution of the exocutebte instmctions causes the processor to perform the following operations:

.10 acquiring target multidimensional feature informatio of a target user wherein foe target multidimensional feature information comprises at feast two type of feature information in biometric feature information and/or nonfoiometric feature information;

comparing the target multidimensional feature information with multidimensional feature informati n of a plurality of designated users, respectively: to obtain a comparison result; a d 15 determining an identity of the target user based on the comparison result,

12. A computer-readable storage medium, storing one or more programs, wherein the one or more programs, when executed by an electronic device comprising a plurali t of appii cations, cause the eleetrome device to perform the following operations:

-o acquiring target muliidimensional feature information of a target user, wherein the target muUidimehsienal feature information comprises at least: two types of feature information in biometric feature information and/or noo-biometric feature information;

2 comparing t!fe large! Multidi ensional feature information vcith multidimensional feature information of a plurality of designated users, respectively to obtain a comparison result, and determining an identi ty of fee _: target user based on fee comparison rssitlt