# US 2022/0044076 A1 · Method and apparatus for updating user image recognition model

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0044076 A1
- Document ID: US20220044076A1
- Status: Published application
- Kind: A1
- Application number: 17/508,401
- Filed: 2021-10-22
- Published: 2022-02-10
- Earliest claimed: 2020-11-30
- Applicants: Apollo Intelligent Connectivity (Beijing) Technology Co., Ltd.
- Inventors: Chuanfei ZOU
- Web page: https://patentgrep.com/patent/US20220044076A1

## CPC

- G06F 18/217: {Validation; Performance evaluation; Active pattern learning techniques}
- G06V 40/172: {Classification, e.g. identification}
- G06F 18/2148: {characterised by the process organisation or structure, e.g. boosting cascade}
- G06F 18/22: Matching criteria, e.g. proximity measures
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/09: Supervised learning
- G06V 10/7788: {the supervisor being a human, e.g. interactive learning with a human teacher}
- G06V 10/94: Hardware or software architectures specially adapted for image or video understanding
- G06V 20/59: inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
- G06V 40/20: Movements or behaviour, e.g. gesture recognition (recognition of facial expressions G06V40/16)
- G06V 40/50: Maintenance of biometric data or enrolment thereof
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/045: Combinations of networks
- G06N 3/08: Learning methods
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A method and an apparatus for updating a user image recognition model are provided. In the method, user image data of at least one target user is acquired; the user image data is input to a pre-trained user image recognition model respectively, to obtain a corresponding recognition result; in responding to determining that the recognition result is an incorrect recognition result, correction information of the incorrect recognition result is acquired; and the user image recognition model is updated based on the user image data corresponding to the incorrect recognition result, and the correction information, to obtain an updated user image recognition model.

## Claims

### Claim 1

A method for updating a user image recognition model, the method comprising:
acquiring user image data of at least one target user;
inputting the user image data of at least one target user to a pre-trained user image recognition model respectively, to obtain a corresponding recognition result, the user image recognition model being configured to recognize the user image data;
acquiring, in responding to determining that the recognition result is an incorrect recognition result, correction information of the incorrect recognition result, the correction information being used to indicate a correct recognition result of the user image data corresponding to the incorrect recognition result; and
updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, to obtain an updated user image recognition model.

### Claim 2 (depends on 1)

The method according to claim 1, wherein the updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, comprises:
updating, in response to acquiring an update instruction containing designated information used to reflect correction information selected by a user, the user image recognition model based on correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information.

### Claim 3 (depends on 2)

The method according to claim 2, wherein the designated information comprises class information of a class to which the correction information selected by the user belongs, and the updating the user image recognition model based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on first correction information and the user image data corresponding to the first correction information, wherein a class of the first correction information matches the designated information.

### Claim 4 (depends on 2)

The method according to claim 2, wherein the designated information comprises a target user identification corresponding to the correction information selected by the user, and the updating the user image recognition model based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on second correction information and the user image data corresponding to the second correction information, wherein a target user identification corresponding to the second correction information matches the designated information.

### Claim 5 (depends on 2)

The method according to claim 2, wherein the designated information comprises a target user identification corresponding to the correction information selected by the user, and class information of a class to which the correction information selected by the user belongs, and the updating the user image recognition model, based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on third correction information and the user image data corresponding to the third correction information, wherein a target user identification corresponding to the third correction information matches the designated information, and a class of the third correction information matches the designated information.

### Claim 6 (depends on 1)

The method according to claim 1, wherein the acquiring user image data of at least one target user, comprises:
acquiring, in response to obtaining authorization of the at least one target user, the user image data of the at least one target user.

### Claim 7 (depends on 1)

The method according to claim 1, wherein the updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, comprises:
updating, in response to a number of a piece of the user image data corresponding to the incorrect recognition result being greater than or equal to a preset number threshold, the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information.

### Claim 8

An electronic device, comprising:
at least one processor; and
a memory communicatively connected with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions, when executed by the at least one processor, cause the at least one processor to perform operations comprising:
acquiring user image data of at least one target user;
inputting the user image data of at least one target user to a pre-trained user image recognition model respectively, to obtain a corresponding recognition result, the user image recognition model being configured to recognize the user image data;
acquiring, in responding to determining that the recognition result is an incorrect recognition result, correction information of the incorrect recognition result, the correction information being used to indicate a correct recognition result of the user image data corresponding to the incorrect recognition result; and
updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, to obtain an updated user image recognition model.

### Claim 9 (depends on 8)

The electronic device according to claim 8, wherein the updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, comprises:
updating, in response to acquiring an update instruction containing designated information used to reflect correction information selected by a user, the user image recognition model based on correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information.

### Claim 10 (depends on 9)

The electronic device according to claim 9, wherein the designated information comprises class information of a class to which the correction information selected by the user belongs, and the updating the user image recognition model based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on first correction information and the user image data corresponding to the first correction information, wherein a class of the first correction information matches the designated information.

### Claim 11 (depends on 9)

The electronic device according to claim 9, wherein the designated information comprises a target user identification corresponding to the correction information selected by the user, and the updating the user image recognition model based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on second correction information and the user image data corresponding to the second correction information, wherein a target user identification corresponding to the second correction information matches the designated information.

### Claim 12 (depends on 9)

The electronic device according to claim 9, wherein the designated information comprises a target user identification corresponding to the correction information selected by the user, and class information of a class to which the correction information selected by the user belongs, and the updating the user image recognition model, based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on third correction information and the user image data corresponding to the third correction information, wherein a target user identification corresponding to the third correction information matches the designated information, and a class of the third correction information matches the designated information.

### Claim 13 (depends on 8)

The electronic device according to claim 8, wherein the acquiring user image data of at least one target user, comprises:
acquiring, in response to obtaining authorization of the at least one target user, the user image data of the at least one target user.

### Claim 14 (depends on 8)

The electronic device according to claim 8, wherein the updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, comprises:
updating, in response to a number of a piece of the user image data corresponding to the incorrect recognition result being greater than or equal to a preset number threshold, the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information.

### Claim 15

A non-transitory computer readable storage medium storing computer instruction, wherein the computer instructions when executed by a computer cause the computer to perform operations comprising:
acquiring user image data of at least one target user;
inputting the user image data of at least one target user to a pre-trained user image recognition model respectively, to obtain a corresponding recognition result, the user image recognition model being configured to recognize the user image data;
acquiring, in responding to determining that the recognition result is an incorrect recognition result, correction information of the incorrect recognition result, the correction information being used to indicate a correct recognition result of the user image data corresponding to the incorrect recognition result; and
updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, to obtain an updated user image recognition model.

### Claim 16 (depends on 15)

The storage medium according to claim 15, wherein the updating the user image recognition model based on the user image data corresponding to the incorrect recognition result, and the correction information, comprises:
updating, in response to acquiring an update instruction containing designated information used to reflect correction information selected by a user, the user image recognition model based on correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information.

### Claim 17 (depends on 16)

The storage medium according to claim 16, wherein the designated information comprises class information of a class to which the correction information selected by the user belongs, and the updating the user image recognition model based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on first correction information and the user image data corresponding to the first correction information, wherein a class of the first correction information matches the designated information.

### Claim 18 (depends on 16)

The storage medium according to claim 16, wherein the designated information comprises a target user identification corresponding to the correction information selected by the user, and the updating the user image recognition model based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on second correction information and the user image data corresponding to the second correction information, wherein a target user identification corresponding to the second correction information matches the designated information.

### Claim 19 (depends on 16)

The storage medium according to claim 16, wherein the designated information comprises a target user identification corresponding to the correction information selected by the user, and class information of a class to which the correction information selected by the user belongs, and the updating the user image recognition model, based on the correction information matching the designated information, and the user image data corresponding to the correction information matching the designated information, comprises:
updating the user image recognition model based on third correction information and the user image data corresponding to the third correction information, wherein a target user identification corresponding to the third correction information matches the designated information, and a class of the third correction information matches the designated information.

### Claim 20 (depends on 15)

The storage medium according to claim 15, wherein the acquiring user image data of at least one target user, comprises:
acquiring, in response to obtaining authorization of the at least one target user, the user image data of the at least one target user.
