Practical Examples of App Inventer Used for AI

lang
en
date
Mar 27, 2024
slug
Post-08-en
status
Published
tags
Tech Share
summary
App Inventor allows developers to use its built-in modules and components.
type
Post
App Inventor is a visual programming environment developed by Google Labs and later transferred to MIT. It uses a building block approach to quickly build applications. With App Inventor, developers can utilize its built-in modules and components, such as multimedia components, drawing and animation components, sensor components, etc., to create applications with AI capabilities.

1:Personal Image Classifier (PIC) Expansion: Train your own AI image recognition model and develop image recognition and classification apps.

 
The basic principle of Artificial Intelligence AI is to prepare the sample data (here refers to the image) and the labeling information of the data (such as the character in the image is happy, angry, crying, etc. image determination information), through the training of AI algorithms, the input samples and the labeling of the fitting, the formation of the final training set of data. With this training set data, when the next time we input a new image, the AI algorithm will be able to determine the specific expression of the character in the picture based on the training set data, so that the picture can be initially classified. Of course, the accuracy of the judgment is related to the number of samples, but also related to the accuracy of the data labeling, but also related to the specific AI algorithm.
PersonalImageClassifier (PIC) Expansion
The events, methods and properties of the extension are as follows:
notion imagenotion image
App development steps
Online AI model training, generate model data, download to the PIC expansion use
Online AI model training website: Personal Image Classifier
Detailed steps of online training: Personal Image Classifier: Part 1
Here are only some of the training steps:
notion imagenotion image
Finally, you can do online AI recognition on new input images, such as smile emoticons:
notion imagenotion image
Finally, you can download the trained model data.
 
App Inventor 2 recognizes and classifies images using the extension and AI model data.
For the usage of PersonalImageClassifier (PIC) extension, please refer to the demo, the documentation is as follows:
 
 
对于本文内容有任何疑问, 可与我联系.