Creating Your First Project
Flexible configuration of the ontology required for the project.
Last updated
Flexible configuration of the ontology required for the project.
Last updated
Before using all the features of the platform, it is recommended to set up your project information in advance, as this can help you quickly configure your data and model objectives in the future.
Hint
Establish a project and set up ontology before beginning any data operations.
Explore the step-by-step process of creating your first AI project. Click on the video below for a quick overview:
On the Project page, you can see a list of all Projects.
Clicking on "Detail" will allow you to view the details, including the specifications and sensor settings of the project.
Clicking on "Create Project" will create a new project. In the popup window, fill in the relevant information as follows:
Step 1: Project Setting
Name: Project Name
Description: Description
Sensor (Stream): The sensor settings and names used in this project, affect the data format of the project. Please choose the sensor combination according to your needs.
Click "Next" to enter the "Ontology" settings for the target object class and details of this project.
Sensor Name
The category and name of your project sensor need to correspond to your annotation format. If you'll be uploading using the visionai format in the future, make sure the data structure and the sensor in visionai.json match what's set here.
Multi-sensor
Multi-sensor ontology is designed to synchronize data from multiple sensors to frame of reference, facilitating temporal alignment and spatial calibration. This approach is crucial when simultaneous data collection from different sensors is needed to construct a comprehensive view of the observed environment.
Selecting a multi-sensor ontology introduces complexity in setup. It generally limits the compatible import formats, primarily requiring adherence to the VisionAI format to ensure proper data preparation.
Step 2: Ontology Setting
Name: Ontology Name
Type: Annotation data format, currently offering the following choices:
Type:
Object Detection (bbox) , Classification, Instant Segmetation, Semantic Segmetation, Point, Polygon, Polyline, VQA
Class: Add the class name that needs to be detected for this project and set a specific color.
Sub-Class: Do classification for this item in Sub-Class.
Attribute: Add or delete attributes flexibly for this class, set the name, type, and field items, and currently have the following types to choose from:
Option: option (vec)
Number: number (double)
Text: text
Boolean: true/false
Ontology Setting
The annotation data format selected in "Ontology" will affect the corresponding dataset and the available AI models.
VQA
When a project ontology type is set to VQA (Visual Question Answering), you can define the question and answer types. This format differs from the standard object detection format.
Step 3: Tag
Tag: For the taggings required for this project, you can flexibly set the name, type, and field items, and currently have the following types to choose from:
Option: option (vec)
Number: number (double)
Text: text
Boolean: true/false
Tags on Ground Truth
The system only recognizes tags present in the ground truth data and does not read tags from other annotation sources. If you require specific tags to be associated with your data and displayed on the platform, please ensure they are included in your ground truth annotations.
Use Dataverse-SDK for Python to help you to interact with the Dataverse platform by Python. Currently, the library supports:
Create Project with your input ontology and sensors
Get Project by project-id
Create Project Example
The create_project
method will create project on the connected site with the defined ontology and sensors.
Please refer to the github content for setting details.
To use the SDK, you need to provide service-id.
When you select SDK in Import Dataset, your Service ID will be displayed in the content, which can be copied and used.