ZOHO AI - Build and Manage Object Detection Model

ZOHO AI - Build and Manage Object Detection Model??

?

?

ZOHO Creator

Microservice

AI Modeler

Object Detection Model

In this page, we will be looking at,

Build Object Detection Model

• Add Training Data - To add datasets for training
• Model Summary

i) Train Model - To train the custom model

iI) Test Model To test the model's performance

Manage Object Detection Model

i) Publish and Use Model - To deploy the model in your applications

?

?

• Object detection model is trained to identify defined objects from digital photographs with a certain accuracy level. Refer to the ?Understand object detection model page to learn more.
• You can??build custom object detection models , suited to your business needs, or choose a??ready to use??model that??can be directly deployed in your applications.

?

To set up an object detection model, follow these 4 steps:

Step 1: Create an object detection model

Step 2: Add training data

Step 3: Verify the model summary, train and test model

Step 4:??Publish and use model

Step 1: Create an object detection model??

1. Navigate to??Microservices under the??DEVELOP section in your Creator??homepage . All your??microservices will be listed here.

1. To create an object detection model :

• If you're creating an AI model for the first time, click the??+ Create New button in the center of the??Microservices page.
• If you've already created AI models, then click the??+ Create New button in the top-right of the??Microservices page.

3. Click the??Create button??in the??AI Models card. The??AI Models home page contains two model types -??Custom models that can be built according to the needs of the user and??Ready-To-Use??models that can immediately be deployed into their applications.

Note : You can create both custom and ready-to-use object detection models. To learn more about the model types, ?refer here .

1. Click the??Object Detection card under the??Build-Custom Models section in the??AI Modeler??page.

1. Enter a??Model Name and click??Create Model. The??Add Training Data screen will appear.Note: The model name cannot exceed??30 characters in length. ?

Step 2: Add training data??

Training data is the initial dataset that is used by the object detection model to analyze data patterns, make interpretations, and arrive at a learning to detect objects. You need to train the model so that it can perceive the input information correctly and make accurate decisions based on the information provided. This ensures that the model performs the way it's intended to. For the ?Object Detection model, a set of images of the object that needs to be detected has to be uploaded as the model training data.

?

1. Click??+ Add New Object??to create an object folder. You can also import images from your .zip folder by clicking??Import from folder .

1. To??help you get started quickly and explore the possibilities of our object detection model, you can download our sample data and start building models.
2. Note : This sample data will be available only when you're creating an object detection model for the first time.
3. Enter the??object's name in the??New Object pop up that appears, then click??Create.??This is the name of the folder where you can add your input images. This object name will be displayed when the model detects your object from the input image.
4. Note : The object folder's name cannot exceed??30 characters in length. If you've uploaded images from your folder, the folder name will be used as the object name. Make sure to give a proper name to your folder before uploading.

1. Double -click the newly-created object folder to add the sample images.

1. Click??Upload Images to add different images of the same object at varying angles.

Note :

1. A??minimum of 10 images must be uploaded for each object.
2. Compatible image formats are JPG, PNG,??and TIF.
3. Each image can have a file size up to 5 MB .
4. For more image guidelines, click??here .
5. Click??Add Images to upload more images. You can also upload multiple images at once by either dragging and dropping them or uploading them as a .zip folder.
6. Compatible image formats for.the zip folder include JPG, PNG, TIF.??

1. You can either click the??Select All radio button or individually select images and click??Delete to remove the images.

1. Click??Done . An object folder will be created.
2. You can add additional objects (folders) by clicking??+ Add New Object??or Import from folder in the dropdown beside it at the top-right corner of the ??Add Training Data??screen

10. Select a folder to??Rename or??Delete it. Click??Select All (ref above image) to??Rename or??Delete all object folders.

11. Click Next . The??Model Summary screen will open.

Step 3: Verify the model summary, train and test model??

After adding the training data, you can review the model details, such as Model Name, Model Type, Model Size, and Total Size. If you need to make any modifications, you can go back to do so. Otherwise, you can proceed to train the model.

• Train model
• View and manage model details
• Test model

Train Model??

Before you can actually use your object detection model in your application, you have to??train it to perform the way you want.

1. Check the details of your model in the ??Model Summary page and make any necessary changes by going??Back . You can modify the Model Name, upload additional images or remove unwanted images.

1. Once you've made the necessary changes, click??Train Model .
2. Note : Model training may take some time, so you can either stay on the page and wait, or you can close the page and come back later.

View and manage model details??

After the training is complete, the user can view the status of the model ( t rained, failed, and draft), the model type, the date it was created on and updated on, and other details as mentioned below.

Model Details??

Under this section, you can view the current version of your model and the names of the added objects.

Version Details??

In this section, you can view the number of versions the model has, what version the model is currently running on, model creation date, the count of objects and their images.

Note : To create a new version of your model, click??here to learn how.

Model Deployment??

In this section, you can view the??App Name ,??Form Name , and the??Field Names in which the model is deployed in. You can also filter between different??environments to check which environment a model is deployed in.

Test Model??

After training, you can test the model's reliability before deploying it in any of your applications. This ensures that the model identifies the test object correctly with a good/high ?confidence score .

??1. Click??Test Model in the top-right corner of the page that appears after the model is successfully trained. This allows you to test the model's accuracy before publishing it.

1. The??Test Model popup will appear. Upload an image of the object that you've added as training data. The model will try to detect the object and display its name on the right side of the popup under ?Model Output .
2. Note : If the model does not detect the test image, you can refer??here to improve your model performance.

After testing your model, you'll get the identified object's name along with a??confidence score .

• If your confidence score is??high, you can proceed to publish your model.
• If your confidence score??is good, you can re-train your model with??additional images .
• If your confidence score??is poor, you need to check for inconsistencies in your training data like insufficient images, make the necessary changes, and train your model again.

Manage Object Detection Model??

After you train your model, you need to publish it to make it available for deployment in your applications.

Note : You cannot un-publish the model once it has been published.You can still make changes to the model and train it again.

• Retrain model
• Publish model
• Use model

Retrain model ??

Retraining the model with the additional images and removing unfavorable images helps your model detect images more precisely. Reworking on the model's efficiency allows the model to be tuned specifically to your business perspective.

Note : We recommend that you periodically re-train your model. This helps in improving the object detection model's reliability and accuracy.

1. Click the??three-dot ellipsis at the top-right corner of your page.

1. Click??Retrain to train your model again. A new version will be created and listed under??Version Details .
2. Click??Edit to make changes to your model. These changes include modifying your model name, object folder name, and adding/removing image. You need to ?train your model again for these changes to be reflected.
3. Click??Rename to edit your model's name. A popup will appear, where you can edit the model's name and click??Rename .
4. Click??Delete to delete your model.
5. Note :

• Deleting a model that is deployed in any of your applications will remove its deployment in those applications. This action cannot be undone.
• After deletion, the added fields ( model input and output fields ) will remain in the form in which the respective model is deployed. All the past data from the object detection model will remain as long as the respective fields are not deleted from the form.
• You cannot delete a model's version that is being current used. Instead, you can switch versions and then delete that model version.

Step 4:??Publish and use model??

Publish model??

After you train and test your model, you can publish it to make it available to your users and start detecting objects.

1. Click??Publish Model in the top-right corner.

1. Click??Publish in the??Publish popup that appears on the screen.

Note: Once a model has been published, you cannot un-publish the same.

?

Use Model??

1. After the model has been published, you can click??Use Model either in the popup that appears or at the top-right corner of your page.

1. Select the??Application Name and??Form Name from the dropdown list in the??Use Model popup that appears, then click??Use Model .
2. The user will be redirected to the form builder of the application they've selected to have the model deployed in.
3. The??Object Detection popup will appear with the??Model Input??screen open.
4. Note: The trained and chosen model will already be selected in the??Select Model??section. You can choose a model when you create a r eady-to-use object detection mode l .
5. In the??Model Input??section, select the source field from the drop down menu and click??NEXT .
6. Note : Currently, you can add only image field as the source field. Therefore, only image type fields available in your form will be listed for source field selection. If there is no image field available in the chosen form, you will need to first create one in order to deploy the object detection model.

1. By default, the model name will be displayed as the field name.You can edit the??Field name and choose a field type ( Single Line or??Multi Line ) for the??Model Output field.

Note : The field name cannot exceed??30 characters in length.

1. Click??ADD FIELD . A new field object detection field will be created.

You can now access your app in live and upload your image which needs to be detected in the??source field. The object detection field will try to detect the image and the image name will be displayed in the ?model output field.