A method, device and electronic device for processing model training labels

Abstract

This application discloses a method, apparatus, and electronic device for processing model training labels. The method includes: obtaining a first image; obtaining object labels for at least one object in the first image, the object labels being used to train a preset teaching training model; outputting the first image containing the object labels, so that the object labels in the first image can be modified; after receiving a label modification operation for the object labels, training the teaching training model using the modified object labels to obtain a new teaching training model; wherein the output parameters of the object in the first image are matched with the modified object labels.

Description

Technical Field

[0001] The present application relates to the field of intelligent education technology, and in particular to a method, device and electronic device for processing model training labels. Background Art

[0002] With the development of technology, the application of artificial intelligence education is becoming more and more extensive. The teaching of training models in artificial intelligence education usually includes four teaching links, such as the annotation of training samples, the feature extraction of training samples, the training of training models, and the testing of training models.

[0003] Currently, only the visualization teaching of the feature extraction phase of training samples and the training phase of training models can be realized, but the visualization of the labeling phase of training samples cannot be realized. Summary of the invention

[0004] In view of this, the present application provides a method, device and electronic device for processing model training labels, as follows:

[0005] A method for processing model training labels, comprising:

[0006] obtaining a first image;

[0007] Obtaining an object label of at least one object in the first image, wherein the object label is used to train a preset teaching training model;

[0008] outputting the first image including the object label so that the object label in the first image can be modified;

[0009] After receiving the label modification operation for the object label, the teaching training model is trained using the modified object label to obtain a new teaching training model;

[0010] The output parameters of the object in the first image match the modified object label.

[0011] The above method, preferably, further comprises:

[0012] obtaining a second image;

[0013] The second image is input into the teaching and training model to output image test data, wherein the image test data includes a test object obtained by processing the second image by the teaching and training model and object attributes of the test object.

[0014] In the above method, preferably, the output parameters of the test object match its object attributes.

[0015] In the above method, preferably, the object in the first image has a label modification mark, and the label modification mark is used to prompt that the object is in a state where label modification can be performed.

[0016] In the above method, preferably, the object in the first image has a label modification control, and the label modification control is used to receive a label modification operation for an object label of the object.

[0017] In the above method, preferably, obtaining an object label of at least one object in the first image comprises:

[0018] performing image recognition on the first image to obtain at least one object in the first image;

[0019] Get the object tag of the object.

[0020] A device for processing model training labels, comprising:

[0021] An image obtaining unit, configured to obtain a first image;

[0022] a label obtaining unit, used to obtain an object label of at least one object in the first image, wherein the object label is used to train a preset teaching training model;

[0023] an image output unit, configured to output the first image including the object label, so that the object label in the first image can be modified;

[0024] An operation receiving unit, configured to receive a label modification operation for the object label;

[0025] A model training unit, used to train the teaching training model using the modified object label to obtain a new teaching training model;

[0026] The output parameters of the object in the first image match the modified object label.

[0027] In the above device, preferably, the image acquisition unit is further used to acquire a second image, wherein the device further comprises:

[0028] The model testing unit is used to input the second image into the teaching and training model to output image testing data, wherein the image testing data includes a test object obtained by processing the second image by the teaching and training model and object attributes of the test object.

[0029] In the above device, preferably, the label obtaining unit is specifically used for: performing image recognition on the first image to obtain at least one object in the first image; and obtaining an object label of the object.

[0030] An electronic device, comprising:

[0031] monitor;

[0032] A processor is used to obtain a first image; obtain an object label of at least one object in the first image, wherein the object label is used to train a preset teaching training model; and output the first image including the object label by the display so that the object label in the first image can be modified;

[0033] The processor is also used to receive a label modification operation for the object label, and train the teaching training model using the modified object label to obtain a new teaching training model;

[0034] The output parameters of the object in the first image match the modified object label.

[0035] It can be seen from the above scheme that in a model training label processing method, device and electronic device provided by the present application, after obtaining the first image and the object labels of the objects contained in the first image, these object labels that can be used to train the teaching training model are output so that these object labels can be modified, and then after receiving the label modification operation for these object labels, the teaching training model can be trained with the modified object labels to obtain a new teaching training model and the output parameters of the objects in the first image correspond to the modified object labels. Therefore, in the present application, by outputting the first image containing the object labels participating in the model training, it is convenient to provide users with a teaching experience in which the object labels participating in the model training can be modified, so that users can intuitively experience the modification of the object labels and the training process of the model, thereby improving the user's learning experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0037] Figure 1 A flowchart of a method for processing model training labels provided in Example 1 of the present application;

[0038] Figure 2-Figure 3 They are respectively exemplary diagrams of the embodiments of the present application;

[0039] Figure 4A partial flow chart of a method for processing model training labels provided in Example 1 of the present application;

[0040] Figure 5-Figure 9 They are respectively another exemplary diagram of the embodiment of the present application;

[0041] Fig.10 A schematic diagram of the structure of a device for processing model training labels provided in Example 2 of the present application;

[0042] Fig.11 Another structural schematic diagram of a device for processing model training labels provided in Embodiment 2 of the present application;

[0043] Fig.12 A schematic diagram of the structure of an electronic device provided in Example 3 of the present application. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0045] refer to Figure 1 , is a flowchart of a method for processing model training labels provided in the first embodiment of the present application, which can be applied to electronic devices capable of image processing and input / output processing, such as computers or servers with input / output devices. The technical solution in this embodiment is mainly used to realize the processing of visualized model training labels to improve the user's learning experience.

[0046] Specifically, the method in this embodiment may include the following steps:

[0047] Step 101: Obtain a first image.

[0048] Among them, the first image can be a sample image participating in the training of the teaching training model. The first image can contain one or more objects, such as flowers, animals, people, weapons and other objects. These objects are used to train the teaching training model so that the teaching training model can perform corresponding processing on the objects contained in the test image, such as object classification.

[0049] Step 102: Obtain an object label of at least one object in the first image.

[0050] In this embodiment, image recognition processing can be performed on the first image to obtain one or more objects in the first image, for example, petal feature recognition is performed on the first image to identify the flower object in the first image, or face feature recognition is performed on the first image to identify the person object in the first image, and so on; then, for these objects, the object label of each object is obtained, for example, any one or any combination of a petal number label, a petal color label, a petal size label, etc. of the flower object is obtained, and for example, any one or any combination of a face shape label, a height label, a body shape label, an action label, etc. of the person object is obtained.

[0051] It should be noted that the object labels here can be used to train the preset teaching and training model. Among them, when the first image is used as a training sample of the teaching and training model, the object labels of the objects contained in the first image can participate in the training of the teaching and training model, so that the model parameters of the teaching and training model are optimized. The teaching and training model here can be a model that can process objects in the test image. For example, after training the teaching and training model constructed based on the classification algorithm using the petal number label, petal color label, petal size label, etc. of the flower object, the teaching and training model can process the objects contained in the test image according to the trained model parameters, and then classify the objects contained in the test image to obtain the object type of each object in the test image, such as whether it is a specific type of flower object.

[0052] Step 103: Output the first image containing the object label, so that the object label in the first image can be modified.

[0053] Among them, in this embodiment, the first image containing the object label can be output through the interactive interface of the output device, so that the user can modify the label parameters of the output object label, such as modifying any one or any multiple of the labels such as the number of petals, petal color, and petal size of the flower object. Therefore, after the object label in the first image is modified, the corresponding output parameters in the corresponding first image match the modified object label.

[0054] For example, the first image includes a flower object, which has a petal number label, a petal color label, and a petal size label. After the first image is output, the user can intuitively understand the object label of the flower object in the first image. At this time, the user can modify the object label of the flower object.

[0055] like Figure 2As shown in , the petal color of the flower object A is changed from the original pink to blue. At the same time, the output color of the flower object A in the first image also changes to blue. The user can intuitively feel the effect of modifying the object label of the flower object A.

[0056] like Figure 3 As shown in , the user changes the petal color of the flower object A from the original pink to blue, and changes the number of petals of the flower object A from the original 4 to 5. The user also changes the petal color of the flower object B from the original pink to green. At the same time, the output color of the flower object A in the first image also changes to blue and the output number of petals is 5, and the output color of the flower object B changes to green. The user can intuitively feel the effect of modifying the object labels of the flower objects.

[0057] Specifically, in this embodiment, the user can modify the object label in the first image outputted by the output device through an input device such as a mouse, a keyboard or a touch screen.

[0058] Step 104: Receive a label modification operation for an object label.

[0059] In this embodiment, a label modification operation for one or more object labels may be received through an interface with an input device.

[0060] It should be noted that the object label corresponding to the label modification operation here can be an object label belonging to the same object, or can also be an object label for different objects. Figure 2 The label modification operation shown in is a modification operation for the petal color label of a flower object A; Figure 3 The label modification operation shown in is a label modification operation for flower object A and flower object B, and is a label modification operation for the petal color label and the petal quantity label of flower object A, as well as a modification operation for the petal color label of flower object B.

[0061] Step 105: Train the teaching training model using the modified object labels to obtain a new teaching training model.

[0062] Among them, in this embodiment, after receiving the label modification operation for the object label, the teaching training model can be modeled using the modified object label to obtain a new teaching training model. As a result, the user can intuitively experience the teaching process in which the modified object label participates in the training of the teaching model, thereby realizing the visual teaching of the model's label modification, thereby improving the user's learning experience.

[0063] It can be seen from the above scheme that in a method for processing model training labels provided in the first embodiment of the present application, after obtaining the first image and the object labels of the objects contained in the first image, these object labels that can be used to train the teaching training model are output so that these object labels can be modified, and then after receiving the label modification operation for these object labels, the teaching training model can be trained with the modified object labels to obtain a new teaching training model and the output parameters of the objects in the first image correspond to the modified object labels. Therefore, in this embodiment, by outputting the first image containing the object labels participating in the model training, it is convenient to provide users with a teaching experience in which the object labels participating in the model training can be modified, so that users can intuitively experience the modification of object labels and the training process of the model, thereby improving the user's learning experience.

[0064] In one implementation, this embodiment can also provide users with a visual learning experience of model testing. Specifically, the method in this embodiment can also include the following steps: Figure 4 As shown in:

[0065] Step 106: Obtain a second image.

[0066] The second image may be an image that needs to be processed, such as a test image in which objects need to be classified, and is used to test the model training effect of the teaching and training model. The second image may contain one or more objects with unknown object attributes, such as flowers of unknown color and size, animals of unknown species, people of unknown identity, weapons of unknown model or number, etc. These objects need to be processed by the teaching and training model to obtain test results such as the objects in the second image and the object attributes of the corresponding objects.

[0067] Step 107: Input the second image into the teaching training model to output image test data.

[0068] The image test data includes a test object and object attributes of the test object obtained by processing the second image by the teaching training model.

[0069] It should be noted that, in this embodiment, after obtaining the second image, the second image containing the unknown object attributes is input into the teaching training model, so that the teaching training model trained with the original and modified object labels performs image learning on the second image through the learned model parameters to obtain each test object in the second image and obtain the object attributes of each test object.

[0070] For example, the petal color label of the flower object A in the first image is pink. Based on this, the training data with the petal color label of the flower object A being pink is used for training in the teaching training model. Afterwards, the new teaching training model obtains one or more test objects X in the second image that have the same or similar object features as the flower object A, and obtains that the petal color of the test object X is pink, such as Figure 5 as shown in .

[0071] Based on this, this embodiment can provide users with a visual learning experience of model testing, and users can intuitively experience the training effect of the model after training.

[0072] Furthermore, step 107 in this embodiment can be executed after step 105. That is to say, combined with the visualization of the modification of the model label in the previous text, in this embodiment, the teaching training model can be trained by the object label modified by the user through the output of the first image, and the second image can be used as a test image to input into the new teaching training model, so that the user can intuitively experience the modification of the model label and experience the training effect of the model trained by the modified label, thereby further improving the user learning experience.

[0073] For example, the user modifies the petal color label of the flower object A in the output first image to blue. Based on this, the teaching training model is trained using the training data in which the petal color label of the flower object A is blue. After that, the new teaching training model obtains one or more test objects X in the second image that have the same or similar object features as the flower object A, and obtains that the petal color of the test object X is blue. Figure 6 as shown in .

[0074] For another example, the user modifies the petal color label of the flower object A in the output first image to blue, modifies the number of petals of the flower object A in the output first image to 5, and modifies the petal color of the flower object B in the output first image to green. Based on this, the teaching training model is trained using the training data that the petal color label of the flower object A is blue, the number of petals of the flower object A is 5, and the petal color of the flower object B is green. Afterwards, the new teaching model obtains one or more test objects C in the second image that have the same or similar object features as the flower object A, and also obtains that the petal color of the test object C is blue and the number of petals is 5. The new teaching model also obtains one or more test objects D that have the same or similar object features as the flower object B, and also obtains that the petal color of the test object D is green. Figure 7 as shown in .

[0075] Optionally, in this embodiment, the second image may be output, and the output parameters of the test object in the output second image match its object attributes. Figure 6 As shown in , the petal color of one or more test objects having the same or similar object features as the flower object A in the second image is blue; Figure 7 As shown in , the petal color of the test object C in the second image is blue and the number of petals is 5, and the petal color of the test object D in the second image is green.

[0076] Based on this, the user can intuitively experience the training effect of the model after training. Furthermore, the user can intuitively experience the training effect of the new model trained with the modified object labels.

[0077] In one implementation, the object in the output first image has a label modification mark, and the label modification mark is used to indicate that the object is in a state where the label can be modified. In a specific implementation, the label modification mark can be marked in the form of a wireframe or an area, such as a square wireframe mark or a circular highlight area, and the label modification mark is output on the first image at an area or position associated with the corresponding object to indicate that the label of the corresponding object can be modified.

[0078] like Figure 8 As shown in , the label modification mark of the flower object A can be represented by a square wireframe. When the mouse hovers over the area corresponding to the flower object A, the square wireframe is highlighted to indicate that the flower object A is in a state where the label can be modified. It should be noted that when the mouse leaves the area corresponding to the flower object A, the square wireframe can be hidden to avoid affecting the user's modification operation on the object labels of other objects on the first image.

[0079] Based on the above implementation, the object in the first image also has a label modification control, and the label modification control is used to receive a label modification operation for the object label of the object. In a specific implementation, the label modification control can be implemented using a menu option, or can be implemented using an input box, so that the user can modify the label of the corresponding object on the label modification control. Therefore, in this embodiment, the label modification operation for the object label can be received through the control interface corresponding to the label modification control.

[0080] like Fig. 9As shown in , when the mouse hovers over the area corresponding to the flower object A, the square wireframe is highlighted to indicate that the flower object A is in a state where the label can be modified, and three label modification controls corresponding to the flower object A are output at the same time, such as the label modification control corresponding to the petal color label, the label modification control corresponding to the petal number label, and the label modification control corresponding to the petal size label. The user can select the petal color parameter in the drop-down menu of the label modification control corresponding to the petal color label, can enter the petal number parameter in the input box of the label modification control corresponding to the petal number label, and can enter the petal size in the input box of the label modification control corresponding to the petal size label. It should be noted that when the mouse leaves the area corresponding to the flower object A, the square wireframe and the label modification controls can be hidden to avoid affecting the user's modification operation on the object labels of other objects on the first image.

[0081] refer to Fig.10 , is a schematic diagram of the structure of a model training label processing device provided in the second embodiment of the present application, which can be applied to electronic devices capable of image processing and input / output processing, such as computers or servers with input / output devices. The technical solution in this embodiment is mainly used to realize the processing of visualized model training labels to improve the user's learning experience.

[0082] Specifically, the device in this embodiment may include the following units:

[0083] An image obtaining unit 1001 is used to obtain a first image;

[0084] A label obtaining unit 1002 is used to obtain an object label of at least one object in the first image, where the object label is used to train a preset teaching training model;

[0085] The image output unit 1003 is used to output a first image containing an object label so that the object label in the first image can be modified; wherein the output parameter of the object in the first image matches the modified object label.

[0086] The operation receiving unit 1004 is used to receive a label modification operation for an object label;

[0087] A model training unit 1005 is used to train the teaching training model using the modified object label to obtain a new teaching training model;

[0088] It can be seen from the above scheme that in a model training label processing device provided in the second embodiment of the present application, after obtaining the first image and the object labels of the objects contained in the first image, these object labels that can be used to train the teaching training model are output so that these object labels can be modified, and then after receiving the label modification operation for these object labels, the teaching training model can be trained with the modified object labels to obtain a new teaching training model and the output parameters of the objects in the first image correspond to the modified object labels. Therefore, in this embodiment, by outputting the first image containing the object labels participating in the model training, it is convenient to provide users with a teaching experience in which the object labels participating in the model training can be modified, so that users can intuitively experience the modification of object labels and the training process of the model, thereby improving the user's learning experience.

[0089] In one implementation, the image acquisition unit 1001 is further used to acquire a second image, wherein the apparatus in this embodiment may further include the following units: Fig.11 As shown in:

[0090] The model testing unit 1006 is used to input the second image into the teaching and training model to output image testing data, wherein the image testing data includes a test object and object attributes of the test object obtained by processing the second image by the teaching and training model, wherein the output parameters of the test object match its object attributes.

[0091] In one implementation, the object in the first image has a label modification mark, and the label modification mark is used to indicate that the object is in a state where label modification can be performed.

[0092] Based on this, the object in the first image also has a label modification control, and the label modification control is used to receive a label modification operation for the object label of the object.

[0093] In one implementation, the label obtaining unit 1002 is specifically configured to: perform image recognition on the first image to obtain at least one object in the first image; and obtain an object label of the object.

[0094] It should be noted that the specific implementation of each unit in this embodiment can refer to the corresponding description in the previous text and will not be described in detail here.

[0095] refer to Fig.12 , is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present application, and the electronic device may be an electronic device capable of image processing and input / output processing, such as a computer or server with input / output devices. The technical solution in this embodiment is mainly used to realize the processing of visualized model training labels to improve the user's learning experience.

[0096] Specifically, the electronic device in this embodiment may include the following structure:

[0097] Display 1201;

[0098] The processor 1202 is used to obtain a first image; obtain an object label of at least one object in the first image, wherein the object label is used to train a preset teaching training model; and the display 1201 outputs the first image including the object label so that the object label in the first image can be modified;

[0099] The processor 1202 is further configured to receive a label modification operation for the object label, and train the teaching training model using the modified object label to obtain a new teaching training model;

[0100] The output parameters of the object in the first image match the modified object label.

[0101] It should be noted that the electronic device in this embodiment may also include a memory for storing applications that can implement the functions of the processor 1202 and data generated by the application being run by the processor 1202. In addition, the memory also includes program content of the teaching and training model.

[0102] The electronic device in this embodiment may also include input components such as a mouse and a keyboard, or the electronic device may integrate an input component on a display, such as a touch screen, etc. The input component in the electronic device can obtain the user's label modification operation and transmit the label modification operation to the processor 1202.

[0103] It can be seen from the above scheme that in an electronic device provided in Embodiment 3 of the present application, after obtaining the first image and obtaining the object labels of the objects contained in the first image, these object labels that can be used to train the teaching training model are output so that these object labels can be modified, and then after receiving the label modification operation for these object labels, the teaching training model can be trained using the modified object labels to obtain a new teaching training model and the output parameters of the objects in the first image correspond to the modified object labels. Therefore, in this embodiment, by outputting the first image containing the object labels participating in the model training, it is convenient to provide the user with a teaching experience in which the object labels participating in the model training can be modified, so that the user can intuitively experience the modification of the object labels and the training process of the model, thereby improving the user's learning experience.

[0104] In the actual application of model training teaching, the technical solution in this embodiment can visualize the label modification and verification test links of the model training samples by providing an interactive scene combined with the artificial intelligence teaching training model, thereby opening up the visualization of the entire artificial intelligence teaching process. Taking the unmanned supermarket model as an example, in this embodiment, the label modification and verification test results in the unmanned supermarket image are visualized to conduct teaching and improve the user's learning experience.

[0105] It should be noted that the images mentioned in the foregoing may be two-dimensional images, as shown in the example figures, or three-dimensional images. The technical solutions formed by different image types or different applicable scenarios are all within the protection scope of this application.

[0106] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0107] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0108] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0109] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for processing model training labels, comprising: Obtaining a first image, where the first image is a sample image participating in the training of the teaching training model; Obtaining an object label of at least one object in the first image, wherein the object label is used to train a preset teaching training model; Outputting the first image including the object label so that the object label in the first image can be modified, after the object label in the first image is modified, the corresponding output parameter in the first image is changed to match the modified object label, so that the user can intuitively feel the modification effect of the object label; wherein the object in the first image has a label modification mark, and the label modification mark is used to prompt that the object is in a state where the label can be modified; the object in the first image has a label modification control, and the label modification control is used to receive a label modification operation for the object label of the object; After receiving the label modification operation for the object label, the teaching training model is trained using the modified object label to obtain a new teaching training model.

2. The method according to claim 1, further comprising: obtaining a second image; The second image is input into the teaching and training model to output image test data, wherein the image test data includes a test object obtained by processing the second image by the teaching and training model and object attributes of the test object.

3. The method according to claim 2, wherein the output parameters of the test object match its object properties.

4. The method according to claim 1 or 2, obtaining an object label of at least one object in the first image, comprising: performing image recognition on the first image to obtain at least one object in the first image; Get the object tag of the object.

5. A processing device for model training labels, comprising: An image acquisition unit, used to acquire a first image, where the first image is a sample image participating in the training of the teaching training model; a label obtaining unit, used to obtain an object label of at least one object in the first image, wherein the object label is used to train a preset teaching training model; an image output unit, configured to output the first image including the object label, so that the object label in the first image can be modified; after the object label in the first image is modified, the corresponding output parameter in the first image is changed to match the modified object label, so that the user can intuitively feel the modification effect of the object label; wherein the object in the first image has a label modification mark, and the label modification mark is used to prompt that the object is in a state where the label can be modified; and the object in the first image has a label modification control, and the label modification control is used to receive a label modification operation for the object label of the object; An operation receiving unit, configured to receive a label modification operation for the object label; The model training unit is used to train the teaching training model using the modified object labels to obtain a new teaching training model.

6. The device according to claim 5, wherein the image acquisition unit is further used to acquire a second image, wherein: The device also includes: The model testing unit is used to input the second image into the teaching and training model to output image testing data, wherein the image testing data includes a test object obtained by processing the second image by the teaching and training model and object attributes of the test object.

7. According to the device of claim 5, the label obtaining unit is specifically used to: perform image recognition on the first image to obtain at least one object in the first image; and obtain an object label of the object.

8. An electronic device, comprising: monitor; A processor is used to obtain a first image, where the first image is a sample image participating in the training of the teaching training model; Obtaining an object label of at least one object in the first image, wherein the object label is used to train a preset teaching training model; The display outputs the first image including the object label, so that the object label in the first image can be modified. After the object label in the first image is modified, the corresponding output parameter in the first image is changed to match the modified object label, so that the user can intuitively feel the modification effect of the object label; wherein the object in the first image has a label modification mark, and the label modification mark is used to prompt that the object is in a state where the label can be modified; The object in the first image has a label modification control, and the label modification control is used to receive a label modification operation for an object label of the object; The processor is also used to receive a label modification operation for the object label, and use the modified object label to train the teaching training model to obtain a new teaching training model.

Images