Appearance

Project Settings ​

General Settings ​

In project General settings, you can configure the basic settings for a project in Labelo.

  1. Project Name: Type the name of your project in the box.
  2. Description: Write a short description about the project in the empty space.
  3. Workspace: Pick a workspace from the dropdown menu to help organize your projects.
  4. Image Description: Select an image description model from the dropdown menu.
  5. Search Method: Choose how you want to search the image description.
  6. Color: Choose a color for your project by clicking on one of the color circles.
  7. Task Sampling: Choose how tasks are selected for the project:
    • Sequential Sampling: Tasks are ordered based on the Data Manager’s preference.
    • Random Sampling: Tasks are selected randomly.
    • Uncertainty Sampling: Tasks are picked based on the model’s uncertainty, which helps in active learning.

Labeling Interface ​

The labeling interface is the main setup area for your project in Labelo. It controls how tasks appear to annotators, ensuring they have the right tools and layout to label data correctly.

It offers a wide range of annotation tools like bounding boxes, polygons, key points, and text labels, depending on the project requirements.

The interface allows users to define custom labeling configurations, making it adaptable to different use cases such as object detection, image segmentation, or text classification.

Models ​

Models in Labelo refer to machine learning models integrated into the platform to assist in automating the annotation process.

Labelo supports both pre-built models and custom models for various tasks such as image classification, object detection, and text analysis. These models can be used for active learning, auto-annotation, auto image captioning and model-assisted labeling, which accelerates the data labeling process by making predictions that annotators can then review and refine.

Note

By leveraging models in Labelo, teams can reduce manual effort, improve annotation quality, and speed up the development of machine learning applications.

Predictions ​

Predictions refer to the automated outputs generated by machine learning models during the annotation process. These predictions help annotators by providing initial labels or insights based on the data being analyzed.. Predictions can be particularly useful in tasks such as object detection, text classification, and image segmentation, enabling faster project turnaround times while maintaining high-quality annotations.

Cloud Storage ​

Cloud storage in Labelo allows users to securely store and manage their data in the cloud, providing scalable and flexible access to project files. This feature enables easy integration with various cloud storage providers, such as AWS S3, Google Cloud Storage, and Azure Blob Storage.

Webhooks ​

A webhook is a mechanism that allows one application to send real-time data to another whenever a specific event occurs.

In Labelo, webhooks enable you to receive notifications about events like project updates or task annotations, allowing you to trigger automated actions in your systems based on these changes. This facilitates seamless integration and efficient workflow management.

Danger Zone ​

The Danger Zone section allows you to perform critical actions that can’t be undone, so use caution. Make sure to back up your data before proceeding.

  • Reset Cache: Use this option if you encounter issues with the labeling configuration or validation errors, and you're certain the labels exist. It helps reset the cache and try again.
  • Drop All Tabs: This option is useful if the Data Manager is not loading correctly. It will close all open tabs related to the Data Manager.
  • Delete Project: Deleting a project will permanently remove all tasks, annotations, and project data from the database. This action cannot be reversed.