Method, apparatus and device for updating feature vector database, and medium
The present application discloses a method, an apparatus, and a device for updating a feature vector database, and a medium. The method includes: acquiring a first biological feature in a service request; obtaining, according to the first biological feature, a first feature vector and a second feature vector respectively through a first algorithm model and a second algorithm model, in which a first feature vector database include sample feature vectors obtained based on the first algorithm model; performing validity verification on the second feature vector according to an associated feature vector for a first user corresponding to a first sample feature vector; and obtaining, under a condition that the validity verification on the second feature vector passes, a second sample feature vector for the first user based on the second feature vector, and storing the second sample feature vector in a second feature vector database.
1. A method for updating a feature vector database, comprising:
acquiring a first biological feature in a service request;
obtaining, according to the first biological feature, a first feature vector and a second feature vector respectively through a first algorithm model and a second algorithm model, wherein a version of the first algorithm model is different from a version of the second algorithm model, and the first algorithm model is configured with a first feature vector database comprising a sample feature vector obtained based on the first algorithm model;
performing validity verification on the second feature vector according to an associated feature vector of a first user corresponding to a first sample feature vector, wherein the first sample feature vector is a sample feature vector successfully matched with the first feature vector in the first feature vector database, and the associated feature vector comprises at least a portion of: the first feature vector, the second feature vector, and at least a portion of history feature vectors corresponding to the first algorithm model; and
obtaining, under a condition that the validity verification on the second feature vector passes, a second sample feature vector of the first user based on the second feature vector, and storing the second sample feature vector in a second feature vector database, wherein the second algorithm model is configured with the second feature vector database.
2. The method according to claim 1 , wherein the validity verification comprises one or more of:
reliability verification, first difference verification, and second difference verification.
3. The method according to claim 2 , wherein the associated feature vector comprises the first feature vector and the history feature vectors corresponding to the first algorithm model; and
the reliability verification comprises:
calculating first similarities between the first feature vector and the at least a portion of history feature vectors corresponding to the first algorithm model respectively; and
determining that the reliability verification passes if none of the first similarities is less than a first judgment threshold.
4. The method according to claim 2 , wherein the associated feature vector comprises the first feature vector and the second feature vector;
the first difference verification comprises:
calculating, using a plurality of second feature vectors obtained based on a plurality of service requests of the first user, a maximum angle among the plurality of second feature vectors;
calculating, using a plurality of first feature vectors obtained based on the plurality of service requests of the first user, a maximum angle among the plurality of first feature vectors; and
determining that the first difference verification passes if the maximum angle among the plurality of second feature vectors is less than the maximum angle among the plurality of first feature vectors.
5. The method according to claim 2 , wherein the associated feature vector comprises the second feature vector;
the second difference verification comprises:
acquiring a plurality of second feature vectors obtained based on a plurality of service requests of the first user;
calculating second similarities between every two of the plurality of second feature vectors respectively;
determining that the second difference verification passes if none of the second similarities is less than a second judgment threshold.
6. The method according to claim 1 , further comprising:
initiating a first auxiliary verification request under a condition that the validity verification on the second feature vector does not pass;
receiving auxiliary information input in response to the first auxiliary verification request;
obtaining an identity verification result by using the auxiliary information for identity verification; and
obtaining, if the identity verification result characterizes that a user indicated by the auxiliary information is the first user, the second sample feature vector of the first user based on the second feature vector, and storing the second sample feature vector in the second feature vector database.
7. The method according to claim 6 , wherein the auxiliary information comprises one or more of a text message verification code, a user terminal token, user identity information and a second biological feature;
a type of the second biological feature is different from a type of the first biological feature.
8. The method according to claim 1 , wherein obtaining the second sample feature vector of the first user based on the second feature vector comprises:
determining the second feature vector as the second sample feature vector; or
calculating a mean vector of a plurality of second feature vectors obtained based on a plurality of service requests of the first user, and determining the mean vector as the second sample feature vector; or
determining a second feature vector acquired last time among the plurality of second feature vectors obtained based on the plurality of service requests of the first user as the second sample feature vector.
9. The method according to claim 1 , wherein after obtaining, according to the first biological feature, the first feature vector and the second feature vector respectively through the first algorithm model and the second algorithm model, the method further comprises:
under a condition that a user of the service request is a known user, and if the first feature vector is successfully matched with a sample feature vector of the known user in the first feature vector database, storing the second feature vector in the second feature vector database as a sample feature vector of the known user in the second feature vector database.
10. The method according to claim 1 , wherein after storing the second sample feature vector in the second feature vector database, the method further comprises:
acquiring a third biological feature in a newly received service request;
obtaining, according to the third biological feature, a first feature vector and a second feature vector corresponding to the third biological feature respectively through the first algorithm model and the second algorithm model;
searching for a third sample feature vector matched with the first feature vector corresponding to the third biological feature in the first feature vector database, and determining a second user corresponding to the third sample feature vector;
performing, under a condition that a sample feature vector of the second user exists in the second feature vector database, matching in the second feature vector database using the second feature vector corresponding to the third biological feature;
determining that the sample feature vector of the second user is valid if a sample feature vector in the second feature vector database successfully matched with the second feature vector corresponding to the third biological feature is the sample feature vector of the second user;
deleting, if none of sample feature vectors in the second feature vector database is successfully matched with the second feature vector corresponding to the third biological feature, or if a sample feature vector in the second feature vector database successfully matched with the second feature vector corresponding to the third biological feature is not the sample feature vector of the second user, the sample feature vector of the second user from the second feature vector database, and continually acquiring a service request comprising a biological feature of the second user to obtain a sample feature vector of the second user stored in the second feature vector database again.
11. The method according to claim 1 , further comprising:
performing, under a condition that a ratio of users having sample feature vectors in the second feature vector database to users having sample feature vectors in the first feature vector database is less than a preset completion coverage rate, biological feature identification for the service request using the first algorithm model and the first feature vector database; and
performing, under a condition that the ratio of the users having the sample feature vectors in the second feature vector database to the users having the sample feature vectors in the first feature vector database is greater than the preset completion coverage rate, biological feature identification for the service request using the second algorithm model and the second feature vector database.
12. The method according to claim 11 , further comprising:
keeping, if a number of second feature vectors does not reach a number required for verification, a sample feature vector of a third user corresponding to the second feature vector in the first feature vector database; and
performing, under a condition that the third user initiates a service request, biological feature identification for the service request using the first algorithm model and the first feature vector database.
13. The method according to claim 11 , further comprising:
performing, if a sample feature vector matched with a second feature vector corresponding to a fourth biological feature in the service request is not found in the second feature vector database, biological feature identification for the service request using the first algorithm model and the first feature vector database according to the fourth biological feature.
14. The method according to claim 13 , wherein after performing biological feature identification for the service request using the first algorithm model and the first feature vector database according to the fourth biological feature, the method further comprises:
initiating a second auxiliary verification request;
receiving auxiliary information input in response to the second auxiliary verification request;
performing, using the auxiliary information, identify verification on a third user obtained by performing biological feature identification according to the fourth biological feature, to obtain an identity verification result;
storing, if the identity verification result characterizes that a user indicated by the auxiliary information is the third user, a third sample feature vector obtained based on the second feature vector corresponding to the fourth feature vector in the second feature vector database, and using the third sample feature vector as a sample feature vector of the third user in the second feature vector database.
15. A device for updating a feature vector database, comprising: a processor and a memory storing computer program instructions; wherein
the computer program instructions, when executed by the processor, implements the method for updating the feature vector database according to claim 1 .
16. A computer-readable storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implements the method for updating the feature vector database according to claim 1 .