Introduction to ExeML
ModelArts ExeML is a customized code-free model development tool that helps users start AI application development from scratch with high flexibility. ExeML automates model design, parameter tuning and training, and model compression and deployment with the labeled data. Developers do not need to develop basic and encoding capabilities, but only to upload data and complete model training and deployment as prompted by ExeML.
Currently, you can use ExeML to quickly create image classification, object detection, predictive analytics, and sound classification models. It can be widely used in industrial, retail, and security fields.
- Image classification identifies a class of objects in images.
- Object detection identifies the position and class of each object in images.
- Predictive analytics classifies or predicts structured data.
- Sound classification classifies and identifies different sounds in the environment.
For the example of building a model using the ExeML function, see Service Developers: Using ExeML to Build Models in the ModelArts Getting Started.
ExeML Usage Process
With ModelArts ExeML, you can develop AI models without coding. You only need to upload data, create a project, label the data, publish training, and deploy the trained model. For details, see Figure 1.
An image classification project detects and classifies images. You can add images and classify them into different classes by labeling. Each class identifies a type of image. After the images are labeled, an image classification model can be quickly generated. It can automatically identify offerings, vehicle types, and defective goods. For example, in the quality check scenario, you can upload a product image, label the picture as qualified or unqualified, and train and deploy a model to inspect product quality.
An object detection project detects whether a certain class of object is included in an input image, labels the position of the object with a proper-sized box, and outputs the detected object class and position. It can identify multiple objects and count the number of objects in a single image, as well as inspect employees' wearables and perform unattended inspection of article placement.
A predictive analytics project is an automated model training application for structured data, which can classify or predict structured data. It can be used for user profile analysis and targeted marketing, as well as predictive maintenance of manufacturing equipment based on the real-time data to identify equipment faults.
A sound classification project identifies whether a certain sound appears in a sound file. It can be used to monitor abnormal sounds in production or security scenarios.