VisionAI Observ
DataVerse
English
English
  • Your Observ Journey
  • Tech Specification
    • Technical Specifications
  • Overview
    • Creating Your First AI Detection Task
    • Managing and Monitoring Detection Tasks
    • Event Records
    • Task Settings
    • AI Detection Streaming
    • Notifications
    • AI Detection Range (ROI)
    • Event Zones (Edit Zone)
    • AI Confidence Score
    • Event Parameters
    • Collecting AI Data
    • Location and Cameras
    • Event Records
    • Event Statistical Analysis
    • Organizations and Users
    • Personal Settings
  • Advance Feature
    • System Settings
    • Exporting Data to DataVerse
    • Managing AI Models
    • Customizing Scenarios
    • Batch Create, Edit, Download
    • Custom Detection Events (Beta)
      • (Event Component) Object Card
      • (Event Component) Area Card
      • (Event Component) Logic Gate
        • Logic Gate: With
        • Logic Gate: Overlap
        • Logic Gate: OR
        • Logic Gate: NOT
        • Logic Gate: AND
        • Logic Gate: Object Tag
        • Logic Gate: Counting
      • (Event Component) Trigger Card
    • Video Task (Beta)
    • Smart Traffic Task (Beta)
    • VLM Playground (Beta)
    • VLM Event-Based Detection (Beta)
  • AI Detection Scenarios
    • Virtual Fencing
    • Smoke and Fire Detection
    • PPE Compliance
    • Pathway Blockage Detection
    • Corrosion Damage Detection
    • Pedestrian and Vehicle Counting
    • Illegal Parking Detection
    • Lane Violation Detection
    • Vehicle Breakdown Detection
    • Customer Counting
    • Queue Monitoring
    • New-PPE and Equipment Compliance
    • Presence and Behavior Monitoring
    • Traffic Violation and Abnormal Behavior Detection
    • Traffic Flow Analysis
    • People Flow and Behavioral Analysis
    • Fire and Smoke Monitoring
    • Water and Pipeline Issues
    • Facility and Access Status
  • API Document
    • API Configuration
    • Event
    • Task
    • Camera
    • Location
    • Organization
    • Compute Target
  • Support
    • Troubleshooting
      • No Image Display Issues
      • AI Detection Issues
      • No Events Detected?
      • Slow or Laggy Live Stream
    • FAQ
  • Observ usage
    • Usage and Billing
  • updates
    • Updates and Release Information
      • Realease 2025/4/10
      • Realease 2025/1/8
      • Release 2024/11/12
      • Release 2024/09/18
      • Release 2024/08/06
      • Release 2024/05/13
Powered by GitBook
On this page
  • Quick Start
  • Previewing and Managing AI Models
  • Adding a New AI Model Repository
  • Importing DataVerse AI Models
  1. Advance Feature

Managing AI Models

The VisionAI Observ allows you to manage current AI models through the AI Model Repository page, import the latest AI models at any time, and keep your detection in optimal condition.

PreviousExporting Data to DataVerseNextCustomizing Scenarios

Last updated 1 month ago

Quick Start

Manage AI models to maximize the space and flexibility of true AI detection. Click on the video below for a quick overview:

Previewing and Managing AI Models

Enter "Detection," switch to "Model," and view the current AI model status. Click "Edit" to make changes.

  • Model Repository: A repository of AI models in multiple versions, with a defined Ontology to determine detection items. AI detection tasks are tied to a Model Repository, but you can switch between different versions of the AI models to match your detection needs.

  • Model: AI models with different version numbers and format types, which can be applied to tasks within the same AI Model Repository, but compatibility with the detection hardware should be considered.

Adding a New AI Model Repository

Click "+ Model Repository" in the upper right corner to add an AI model. Fill in the following fields:

You can choose to add a new AI Model Repository or copy an existing one.

  • Repository Name: Enter the name of the new AI Model Repository.

  • Repository Description: Provide a description of the AI Model Repository.

Importing DataVerse AI Models

Click "+ Model" in the upper right to import AI models from DataVerse.

You can choose to add a new AI Model Repository or import models into an existing one.

  • Repository Name: Enter the name of the new AI Model Repository.

  • Repository Description: Provide a description of the AI Model Repository.

  • Dataverse Project Model: Select the project in DataVerse and the corresponding AI model, which relates to the target items of this AI detection.

  • Model Description: Describe the AI model.

Click "Import" to import the AI model.

Once you've established a new AI model using the above steps, you can choose this model for use in the event editor.

Tip: When you import a new AI model, remember to check the system settings to confirm if you want to replace this new model for task detection.

You can manage current AI models through the AI Model Repository page, import newly trained models at any time, and keep your detection in optimal condition.

How to visit Dataverse? Start using it here 👉

https://linkervision.gitbook.io/dataverse/
System Settings